Computer Science Research on Smart Grid Information System Architecture Cao Junwei) 2) Wan Yuxin 3 Tu Guojun Zhang Shuqing 4) Xia Ai 3 Liu Xiaofei Chen Zhen 12 Lu Chao 4) Tsinghua University Institute of Information Technology, Beijing 100084) 2) (Tsinghua Trust Science With technology countries.

The smart grid mainly solves the following problems: ensuring grid security, stability and reliability while improving equipment utilization. Due to the high coupling of the power grid system and improper dispatch control, a single fault can cause cascading failures, and even cause large-scale power outages and equipment damage, resulting in inestimable direct and indirect losses. Therefore, the reliability requirements of the grid system are very high. The intelligent scheduling of the smart grid is to solve the problem of collecting, transmitting, analyzing and processing wide-area information on the basis of ensuring security and reliability.

Realize the interaction between power generation and electricity use. The basic feature of the grid is the balance between power generation and electricity use. From the perspective of the network user, the user can obtain the operating parameters of the power grid (such as the cost of power and the power consumption of various devices) through the intelligent power network, thereby adjusting the power usage of the power. For the power grid system, an accurate load model can be constructed according to the power consumption information of the power equipment, thereby effectively improving the power supply efficiency. The construction of the traditional power grid is based on the one-way thinking of the transmission-transmission f, and a large amount of redundancy causes waste. The smart grid is based on a high-real-time (tens of milliseconds) measurement communication system, which can achieve power generation load balance through real-time control. Thereby, the hot spare can be reduced and the stability of the system can be improved.

Intermittent renewable energy access. New energy mainly refers to wind power and photovoltaic power generation. China's wind power resources are mainly concentrated in the northwestern region, and these areas are also areas with abundant solar energy resources.

However, China's power demand is concentrated in the central and eastern regions, so China's new energy power must be transmitted long distances to reach the load zone. This requires the grid to be optimally configured for new energy generation across the country. At the same time, due to the randomness and intermittent nature of new energy power generation, if it is directly integrated into the power grid, it may affect the overall stability of the power grid system. If wind power generation may be disconnected from the network due to objective meteorological reasons, it will cause instantaneous imbalance of the power system, which in turn will affect the overall stability.

It can be seen that the smart grid needs to solve the bottlenecks of traditional grid information systems in information collection, transmission, processing and sharing, and the solution of these problems depends on the evolving Internet of Things technology. The core technology of the Internet of Things covers physical state perception, information representation, information transmission and information processing from the sensor network to the upper application system. It will play an important role in communication, security and upper-layer applications in the smart grid information system. Function: Sensor network technology can be used for data acquisition and information acquisition of communication devices such as smart meters; real-time and secure communication technologies can be used for transmission of grid operating parameters, real-time transmission of grid operation and maintenance data and power generation load data; data The storage and information representation technology 6 can be used for storage, management, query and organization of massive data of the power grid; the data distributed processing and the arbitrary scheduling technology can be used for power system security stability analysis, real-time deployment of energy flow after new energy access. The development of the Internet of Things technology has enabled the power system to be integrated into the computer digital environment from a relatively phase 1 Cao Junwei, etc.: Intelligent grid multiplication system architecture research, closed self-sufficient control system, while improving the stability of the grid, making wind, electric energy, etc. Energy is easily integrated into the smart grid information system for unified planning and scheduling.

Drawing on the information technology architecture of the Internet of Things, this paper proposes the architecture of the smart grid from the perspective of information technology. Section 2 summarizes the definition of smart grid; Section 3 introduces the current status of smart grid development at home and abroad, major technical difficulties and challenges, and proposes a smart grid information system architecture; Section 4, Section 6 details the infrastructure of the Smart Grid Information System, Support platform and application system; Section 7 summarizes and looks forward to the future research direction of smart grid.

2 Smart grids define smart grids, usually referred to as modern grid systems that incorporate modern information systems into traditional grid networks. Therefore, the power grid has better controllability and observability, and solves the problems of low energy utilization, poor interaction, and difficulty in analyzing safety and stability of traditional power systems. At the same time, based on real-time regulation of energy flow, it is convenient for distributed new energy generation and distribution. Access and use of energy storage systems.

The first significant feature of the smart grid is its considerable performance. That is, by means of information network technology, real-time monitoring of information of each node of the power system.

For example, IBM defines the first level of the three levels of smart grid as real-time, comprehensive and detailed monitoring of information, eliminating blind spots in monitoring. “Tsinghua University proposed “CCCP” in the 1980s (communication, computer and control technology in the power system) Applying the concept, the smart grid is the integration and interaction of the two networks of the traditional power system network and the power information network. The same definition also includes related concepts.

Another feature of the smart grid is the dynamic interaction between the two sides of the power generation. That is to use the real-time acquisition of grid power generation information and user information for optimal scheduling. From the point of view of the network users, the goal of the smart grid is to co-ordinate all power resources to provide more stable power to the network users in a cheaper way. For example, Duke Energy proposed that in the smart grid environment, network users can observe their own power consumption in real time and adjust their own electricity habits to reduce costs. At the same time, power companies can deploy according to users' needs. Supply and price guidance to guide users' needs, reducing total energy consumption. The European Smart Grid Strategic Development Plan proposes that the smart grid should integrate all the users, generators and two-way equipment connected to the grid, and strengthen the control of the power generation side through intelligent monitoring, communication and self-healing technologies, and provide users with more information. And electricity optimization programs to reduce the impact of the power system on the environment, improve the reliability and safety of power supply.

The third feature of the smart grid is its high reliability. That is, it can be automatically recovered from the system oscillation, and the system is alarmed and adjusted in advance for the system instability trend. For example, the US Department of Energy defines smart grids with features such as system oscillation self-recovery, high robustness, and security. The third level of the three levels of smart grid defined by IBM is to conduct advanced analysis based on information integration to achieve the goal of improving reliability, reducing costs, and improving revenue and efficiency.

Based on the above viewpoints, we define the smart grid as follows: Smart grid is a comprehensive composite system integrating sensing, communication, calculation, decision-making and control built on the basis of traditional power grid. The operation status of the node resources and equipment, hierarchical management and power allocation, to achieve a high degree of integration of energy flow, information flow and industrial labor, improve the operational stability of the power system, in order to maximize the efficiency of equipment utilization, Improve safety and reliability, save energy and reduce emissions, improve the quality of power supply for users, and improve the utilization efficiency of renewable energy. The most important goal of the smart grid is to reduce energy consumption costs, improve the quality of residential electricity consumption, and reduce the operating cost of electricity, thereby promoting the development of the national economy.

3 Status and Challenges of Smart Grid Development 3.1 Development Status of Smart Grid at Home and Abroad In 2001, the CIN/SI project was launched, and a modeling, simulation, analysis and synthesis tool was developed to establish high robustness, high adaptability and control. Reconstructed networked power systems and infrastructure, introduced in June 2001 by "Wred", is an earlier reference to the construction of smart energy networks. After that, the American Electric Power Research Institute launched the IntelUGrul project and released the IntelUGml1 architecture in 2004. GE, Cisco, and Lucent participated in the research and development of the project. The project aims to integrate energy systems and control information systems in power systems, and to provide implementation steps and technical guidance on how to build smart grids from the perspectives of power information systems and service models. In 2003, the US Department of Energy released the grid 2030 Blue Dragonfly and in the same year led the establishment of the GridWise2 Alliance, which aims to promote the integration of traditional power systems and information technology to build a new smart grid. At present, members include IBM, Sco, West Gate, GE, Microsoft, Samsung and more than 140 companies in the energy and information fields. In March 2008, Xcel Energy Corporation of the United States announced the establishment of a smart grid city pilot project in Boulder, Colorado. Currently, 23,000 intelligent monitoring devices have been installed to provide users with more convenient and stable power supply and help users save money. Electricity costs. In May 2011, the United States established a new smart grid pilot in Maui, Hawaii. In general, the development of smart grid technology in the United States focuses on the integration of communication technology, control technology and power system, while emphasizing the interaction between network users and grid systems. The US Department of Energy's Smart Grid Report 2009 indicates that building a smart grid system should be carried out in six areas: transmission systems, distributed energy, power distribution systems, information networks, management, and financial environments.

The construction plan of the European smart grid began in 2004. At the first International Conference on Renewable Energy and Distributed Energy Integration, industry and research stakeholders proposed the idea of ​​establishing a future European power network technology platform. In 2005, with the support of the European Commission, Europe established the Smart Grid European Technology Platform to provide planning for the development of European power networks in 2020 and beyond. The organization released the European Smart Grid Design Blueprint in 2006, proposing that the smart grid must include four objectives: flexibility, accessibility, reliability and economy. The accessibility section specifically mentions renewable energy and Access to efficient low-carbon capacity. At the end of 2008, the organization released the European Smart Grid Strategic Development Plan and released the most comprehensive version in April 2010. The European smart grid development was prioritized into six levels, covering grid optimization, distributed energy, and information and communication technologies. To market operations and other aspects. All targets will be completed around 2020, with the first phase of the goal (optimizing grid operations and usage) in 2008 2012 to address grid operation, safety and market-oriented energy flow control issues in a distributed environment. In April 2009, Sakamoto announced the “Sakamoto Development Strategy and Economic Long-Term Plan”, which includes solar power grid-connected, future smart grid power grid test, and electric vehicle fast charging device, which are closely related to the smart grid. The Sakamoto Electric Business Association said in July 2009 that it will fully develop the "Spirit Edition Smart Grid". South Korea released the "Green Energy Industry Strategy" in 2008 and launched the "Korean version of the Smart Grid".

Imagine.

In China, in May 2009, the State Grid Corporation of China proposed the development plan of China's smart grid, which will promote the construction of China's smart grid in three phases and plan to build a unified smart grid in 2020; from the grid itself to security and stability, energy scheduling, Eight aspects of user interaction and new power applications give the characteristics of China's future smart grid. North China Power Grid Corporation announced the pilot community in April 2009. In 2008, China Southern Power Grid Corporation became a wide-area damping control system, which is the world's first wide-area closed-loop intelligent control system that has been successfully implemented. The system is based on wide-area closed-loop control that is adaptive by the synchronized phasor measurement unit. As of 2009, China has imported more than 1,000 PMU nodes in Northeast China, North China, Central China, Jiangsu, East China, Henan, Yunnan, Lizhou, Guangdong and South China, and established more than 10 WAMS central stations, covering 500kV substations. And major power plants, more than the United States in this regard.

There are also a number of units in the university that are conducting research on smart grids. For example, the team led by Academician Han Yingjun of Tsinghua University has achieved certain results in the monitoring of wide-area power grids, and solved the key technical problems in the wide-area damping control project, and proposed cooperation with Sifang Group. A power system security early warning, defense and control system construction scheme based on wide area information. The team led by Academician Yu Xinxin of Tianjin University has made certain progress in the distributed generation function system, and put forward the technical idea of ​​connecting the distributed generation function system of solar energy, wind energy and small water energy into the large power grid in the form of microgrid. Improve energy transmission efficiency and power transmission stability and reliability, improve power quality and reduce costs.

Overall, the current research on smart grids focuses on the access of distributed energy and the interaction of power generation. The research work of smart grids in China focuses on the information acquisition and stability control of large grid systems, which is compatible with China's power network. Strong characteristics are related.

3.2 Status and main problems of grid information system The existing grid information system (power secondary system) mainly refers to the power dispatching automation network and its energy management system EMS (Energy Management System), distribution network management system DMS (Distribution Management System) and wide area monitoring System WAMS. Energy management system mainly includes data acquisition (Acquisition), automatic power generation control system AGC (AutomaticGamControl) and power state estimation system; distribution network management system mainly includes distribution automation system DAS (Dstrbuion Information System) and demand side management system DSM ( DemandSideManagement), etc.; and the wide-area monitoring system consists of a synchronous phase angle measuring unit PMU, real-time acquisition of the main data of the grid. The EMS and DMS systems all rely on the remote control unit RTU (RemoteTerminalUnit) and its data acquisition and monitoring system SCADA. The main problem is that the data acquisition time is too long, reaching the sub-second level, which cannot meet the real-time requirements of high-efficiency applications such as the power grid. Wide area control, energy scheduling, etc. WAMS Phase 1 Cao Junwei et al.: The response time of the smart grid multiplication system architecture research system is on the order of 100 milliseconds, but the WAMS system relies on the construction of the power private network, and the input cost is relatively high. Currently there is no PMU below the domestic 110kV voltage level. Node deployment. In addition, the existing grid information system only collects and controls data for the distribution of electric field stations and high-power electrical equipment, and cannot obtain real-time information of the load. The energy allocation is also based on offline prediction. This has caused four major problems facing the existing power network: (1) the important parameters of the power system are random, time-varying, and unobtrusive, which makes the power system prediction and scheduling difficult; (2) the true electrode limit of the transmission line is unknown, often Relying on great conservativeness for reliability, resulting in low line utilization; (3) for faults in long-distance transmission, it is impossible to accurately know the fault information, such as the location and severity of the fault, often adopting a tentative approach to deal with the fault, resulting in a large amount of equipment redundancy; (4) The power system cannot be stored actively, the reactive power cannot be dynamically balanced, the load cannot interact, and the hot standby causes waste.

In order to solve the above problems, a large number of sensing devices, such as smart meters and PMU units, need to be added, and the addition of sensing devices means that the amount of real-time data is large, and the real-time transmission and processing of power system data under large data volume needs to be utilized. Advanced information, communication, networking and computing technologies are the issues that smart grid information systems need to address. Based on this, we propose the following structure of the smart grid information system.

3.3 Smart Grid Information System Architecture The proposed smart grid information system architecture, as shown in Figure 2, mainly includes three parts: the smart grid information system infrastructure, the smart grid information system support platform and the smart grid information system application system.

Schematic diagram of smart grid information system architecture Smart grid information system infrastructure mainly refers to the hardware foundation of building smart grid, while smart grid information system support platform mainly refers to the software infrastructure of building smart grid, on top of which is to achieve the goal of smart grid construction Various types of applications. The above three platforms will be explained separately below.

4 Smart Grid Information System Infrastructure The smart grid information system infrastructure is the hardware foundation for building a smart grid, including the main links of the power system and the control, measurement equipment and communication network.

4 1 Power system control and measurement equipment First, the composition of the power system is briefly introduced. The power system is mainly composed of six parts: power generation, transmission, power transformation, power distribution, power consumption and dispatch. Power generation includes traditional hydropower, thermal power and new nuclear energy, wind energy and solar power generation. The control of power generation is mainly for frequency regulation of generators, voltage amplitude adjustment, synchronous phase and active reactive power regulation, and generator output. The voltage is generally in the range of 1135 kV. The transmission link connects the main generators and load centers in the grid system to form the backbone network of the grid system, usually operating at the highest voltage level (eg 220kV or more). Commonly used transmission technologies include high-voltage direct current transmission and flexible alternating current transmission. The substation link becomes a secondary distribution process of power, connecting substations and substations, and some large industrial loads may directly access the substation system.

The voltage level of the substation system is generally between 69138 kV. Through the transformer ratio and reactive power compensation equipment for the substation, the grid system can control the reactive power and voltage of the grid. The distribution link is the most integrated into the conversion of electric energy to individual users. The distribution system is divided into a primary distribution system and a secondary distribution system. The primary distribution system mainly supplies small industrial electricity, and the voltage level is 434 secondary. The electrical system is used for electricity for residents and businesses, with a voltage rating of 120240V.

The power system measurement equipment is the basis for building a smart grid. The implementation of the smart grid depends on the application and deployment of the sensor. Currently, the sensors in the smart grid include the grid operation and maintenance measurement system and the personal user measurement system. The grid operation and maintenance measurement system is mainly used to collect electrical information of power system units such as transmission and distribution lines, power plants, and motor sides. Commonly used remote network devices such as SCADA systems, RTUs and PMUs in the WAMS system have measurements. , communication, control and other functions, the measurement unit is widely used in energy management systems (EMS), but its main disadvantage is that the data sampling frequency is low, can not get the dynamic information of the grid operation in time; each RTU unit is not synchronized The clock and the acquired data are out of sync. Compared with the RTU unit, the PMU adds phase angle measurement; it has a GPS timing unit for higher measurement accuracy; at the same time, the measurement frequency is higher, on the order of tens of milliseconds. The personal user measurement system is mainly used to measure personal power usage, such as smart meters. The main function of the smart meter (SmartMeter) is to obtain the power-saving and energy-saving suggestions for the user by obtaining the power consumption data of different power-consuming devices of the user and combining with the operation of the power grid, and the information flow is transmitted in both directions. Smart meters should have the following functions: two-way communication; automatic data collection; power outage management; dynamic billing management; demand response for load control.

At present, in the field of smart meter development, there are two main ideas: (1) using multiple acquisition devices to directly collect data from the appliance; 2) using a collection device to collect data, and then using a classification algorithm to identify the data. One of the drawbacks of the first idea is that each appliance needs to be equipped with sensing equipment, which is costly, and some electrical appliances are difficult to install and require additional communication protocols and equipment to support data collection. Relatively speaking, the cost of the second idea is relatively low, mainly based on the pattern recognition algorithm to classify the electrical characteristics of the electrical appliances, and thus analyze the power consumption of different electrical appliances. For the second method, it is important to use the sensor to obtain which type of data, because the selection of features has a great influence on the efficiency of the pattern recognition algorithm. At present, the existing design has the advantages and disadvantages of the measurement scheme of Table 1. The difference between the active power and the reactive power and high-power electrical appliances is large, which is advantageous for distinguishing (1) not suitable for small-power appliances (2) similar electrical appliances cannot be distinguished (3) implementation is complicated, It is necessary to simultaneously measure the power and reactive power and current characteristics and start-up characteristics. It is easy to distinguish similar electrical appliances. (1) Inaccurate measurement of high-current electrical appliances. (2) Professionals need to install transient voltage and noise characteristics. (1) Easy installation, any socket Can be used for installation () can be used to distinguish similar appliances (1) each user needs to retrain (2) requires high adoption rate (above MHz) to measure continuous high frequency voltage characteristics requires higher sampling frequency (50 power control equipment is realized The carrier of the smart grid target, the main operating parameters of the grid system are frequency, voltage, phase, active power, reactive power. To achieve the control of the above parameters, the control object of the grid system includes all levels of power generation units, power transmission and transformation systems, Power distribution system. The main control equipment is RTU unit and various intelligent electronic equipment (IntelligentElectronicDevice

4.2 Power System Communication Network Communication network is an important infrastructure for smart grids. The wide-area measurement system WAMS, the Wide Area Protection System (WAPS), and the Wide Area Control System WACS (Wide Area Control System) in the smart grid all depend on the communication architecture. Due to the diversity and dispersion of the grid system, there is no unified system architecture for the grid system. Considering the current networking mode and future application requirements of the smart grid, we believe that the bottom-up architecture of the smart grid communication network system is as shown in Figure 3.

According to the smart grid communication network system, the smart grid communication network can also be viewed as two parts: 1) The power state monitoring network consisting of the grid state measurement unit PMU and RTU. The number of nodes in the domain is small. (2) An information network composed of individual user measurement units, which is characterized by a large number of nodes and high scalability requirements.

4 2.1 Personal User Network The personal user measurement unit often connects through the local area network and then accesses the WAN. The local area network composed of the smart meter connection includes the home area network HAN (HomeAreaNetwork) and the available networking modes include a wireless network and a BPL (Broadband OverPowerLine) network. Among them, the wireless network is used to construct a smart grid personal user LAN. The existing standards include the Zgbee1 protocol and the OpenHAN protocol. Both of the above protocols operate on the basis of the IEEE 802. 15.4 wireless network standard. Zgbee protocol is a common networking technology in wireless sensor networks. It is mostly used in the construction of low-speed short-range wireless networks. A scheme based on Zgbee to construct personal user LAN is given. OpenHAN is a wireless network networking protocol designed specifically for home power systems. In 2008, the first edition of the networking requirements document was released by the Open Smart Grid User Group (OSGUG) (OpenSmartGridUsersGroup) and revised in 2010. The networking structure for constructing a personal user LAN is either a star network or a mesh network. The main disadvantage of the star network is that the central node has a heavy burden and there is a single point of failure problem. The mesh network is more common in the structure of the wireless sensor network. Due to its better self-healing characteristics, the mesh network is actually used to construct the phase one. Cao Junwei et al.: The smart grid multiplication system architecture studies the individual user power information network, but the nodes near the centralized node (AccessPmnt) are the bottleneck of the mesh network.

42.2 Power backbone communication network The smart grid backbone communication network networking mode can be divided into two categories. The first type is the combination of power network and information network, that is, the communication carrier itself is an element in the power network, including based on GroundWre) Media self-supporting overhead cable ADSS (A11DielectricSelfSupporting). The second category is the separation of the architecture of the smart grid information network from the power network, that is, the use of an additional network architecture power system information network. In this mode, there are also different information network architectures, which can be roughly divided into three types, namely, optical fiber, wireless signal, and leased band competition. At present, the common practice is that the backbone network is constructed by fiber, and the edge network is transmitted by wireless.

The use of power network elements to construct an information network model is conducive to cost savings, but it is easy to cause the power system and information system to be coupled to each other. The failure of the power network will lead to the failure of the information network. The separation mode can solve the above problems and make the smart grid information network architecture more free. However, in the separation mode, the information network must separately select the transmission carrier, which needs to balance the cost and transmission performance. In particular, power system equipment has a wide distribution range. Some remote areas do not have the conditions for constructing optical fiber or wireless networks, and require additional transmission methods. For example, based on the transmission architecture model of cognitive radio CRCCogmtveRado, the benefit of cognitive radio lies in Find the white space suitable for communication from the frequency band of a specific area, and use the transmission belt to compete without affecting the existing communication system. The IEEE 802.22 protocol defines a white space search method. At present, the 802.22 protocol has been deployed in video video, and it can be used to construct information networks in remote areas through CR technology.

The existing power system network communication protocols include the IEC60870, IEC61850, and IEC61970 protocol groups. Since the above protocol groups are mainly constructed for different types of data networks, this section will not be described, and will be further described in Section 6.

42.3 Main Indicators of Smart Grid Communication Network The two main indicators for the construction of smart grid communication network are network stability and network delay. Different network construction methods will inevitably lead to different network characteristics. How to select the construction scheme of smart grid communication network is an important issue in the field of smart grid research.

There are two ideas for discussing the delay and stability of smart grid networks: (1) From the perspective of network topology and protocol itself, such as the study of the network with dedicated competition and sharing competition in the separation information network architecture mode. Performance and its influencing factors; 2) Study the transmission performance of smart grid from the perspective of information theory, such as the analysis of the channel capacity required for smart grid wireless communication to ensure the requirements of secure communication. On the basis of network performance analysis, considering the impact of power system communication delay on control performance is a problem that networked control systems need to solve. However, current networked control systems usually analyze for a single network and are not popular.

As mentioned above, the current power information network is usually built by a private network. However, due to the cost constraints of dedicated deployment, the dedicated band competition is often not very large, and in this case, the shared band competition model can often obtain a larger channel. Capacity, which means better transmission delay performance, but the problem is that the stability of the delay under the shared-band model is greatly affected by network conditions. If the background noise ratio is high, the network delay and packet will rise rapidly. . Based on TCP/IP, the smart grid WAMS and WAMC models were built, and the network delay and packet conditions after adding background noise and QoS mechanism in the shared band competition mode were analyzed. At the same time, the situation of different competitions under the shared competition model was compared. At present, China's smart grid adopts a private network construction mode, but due to cost constraints, it is limited to 220kV and above. How to ensure the real-time and stability of transmission in the case of sharing a competitive network will be a difficult point.

For the personal user measurement network, it is characterized by a large number of network nodes, but the amount of data of a single node is limited. In addition, the network is mostly built by wireless. How to collect these data and ensure the real-time performance of the data is needed for the smart grid. problem. A smart meter measurement system model based on compression sensing technology is proposed for the personal user measurement system. The wireless access method is adopted. On the other hand, with the increasing number of wide-area monitoring nodes, the existing power information The network is gradually unable to meet the system requirements, and the pressure of the large data volume on the competition is also likely to cause delays. If the power system raw data itself can be compressed, the system can reduce the need for competition. A measurement system model based on matrix singular value decomposition is proposed for grid operation and maintenance measurement system. By analyzing the degree of coupling of the grid connection to determine which data needs to be transmitted between the areas, the size of the data to be transmitted is reduced.

The main difference between smart grid communication transmission and traditional information transmission is that the dynamic of the system is strong. The difficulty of smart grid communication lies in its high stability requirement for network delay and time delay. The main problem faced by traditional power grid communication control systems, such as SCADA systems, is that the delay is too large. How to balance the cost and performance according to the limitations of physical conditions is also a difficult problem for the main computer journal of smart grid research in the future. In addition, how to ensure the confidentiality and security of the smart grid data channel is also a problem to be solved.

424 Smart grid upper-layer application network With the promotion of distributed generation and energy storage technology, from the perspective of power supply and use, the self-organization characteristics of the grid will be strengthened, and the grid will exhibit self-production and sales characteristics within the local area. For example, the power used by users in the future may be partly from the supply of large power grids, while the other part comes from the power generated by new energy sources in its vicinity. The loss of transmission and distribution in this mode will be reduced and will help to reduce the load on the large grid. The network model of this self-organized power supply network is consistent with the content distribution network CDN (ContentDeliveryNetwork). The power supply mode similar to Internet Cache and P2P may also be generated in the power grid, that is, through the hybrid vehicle PHEV (PlugrinHybrid ElectricVehicle) and electric The car EV came to act as a Cache, and a similar idea was put forward in 2004. This paper believes that the upper-layer application network of the smart grid can be constructed by using overlay network technology and information center network technology. ) is a virtual network based on current TCP/IP architecture Internet communication. It improves communication reliability and service quality on TCP/IP networks by deploying a set of nodes on the existing communication infrastructure. Coverage networks provide the foundation for network communications for the diversity of smart grid applications. For example, the load balancing problem in the microgrid system can be solved by the P2P model. With the algorithm of P2P technology in distributed resource discovery, the system can quickly acquire the data of power consumption and power generation of each node, and then perform the deployment. If regional information such as address is added to the node description, the system can reduce transmission loss according to the principle of nearby supply. An agent-based microgrid transmission and distribution deployment model is described. In addition, P2P technology can be applied in many aspects such as power pricing system, intelligent protection system, and intelligent unloading. Coverage network technology can also be used to improve the security and latency performance of smart grids. By setting up a security hub (hub), the data concentrator can select a secure data forwarding node by means of authentication or the like. And the use of overlay network technology can help improve the overall reliability of the network, and should not cause single point failure.

(InformationCentricNetwor-kmg, ICN) is one of the important achievements of the current future Internet architecture research. The basic idea is to separate the information object from the network location, and to apply the early warning analysis based on the important role of the publish/subscribe model. The challenge of communication and security. In addition, grid state measurement is often directed to large grid components or loads. The number of local area is relatively limited and the performance of the measurement unit is high. However, the system often adopts a centralized management mode, which results in excessive load on the central node of the system. The competitive constraints are obvious.

Since the 1990s, the intelligent meter reading equipment (AMR) has gradually begun to apply the pilot, but AMR has only become a remote acquisition and billing function of data, and does not have the function of regulating the user's power consumption behavior. Pass to pass. The Advanced Metering Infrastructure (AMI), which is composed of Smart Meters (SMM), can realize the two-way transmission of information flow. The smart meter and AMI system are the basis for building a smart grid. Compared with the grid state measurement, the personal measurement system is characterized by a large number in a small area and high scalability requirements; at the same time, it requires real-time data and security.

The smart grid measurement system is the basis for the realization of the smart grid, and the power data collection function is realized. Existing measurement systems include SCADA systems, WAMS systems, and AMI systems. The SCADA system and the WAMS system are combined to collect power state data and the AMI is collected into individual user data. The SCA-DA system is not very real-time and is gradually being replaced by the WAMS system. The AMI is still in development and has not yet formed a forming plan. On the other hand, the role of the smart grid measurement system depends on the data analysis processing system. The smart grid data representation and storage architecture will be analyzed below.

5.2 Data Representation and Storage System 52.1 Smart Grid Data Representation Since the grid system equipment is jointly produced by a number of different manufacturers, how to describe the grid system itself and uniformly manage the data generated by these heterogeneous devices is the key to realizing the smart grid information network. one. The representation of the power grid system includes the naming of the data collected by the power system, the definition of the data, the description of the device, the representation of the relationship between the devices, and the representation of the communication model.同样,智能电网的数据表示可以划分为电力系统数据表示和个人用户数据表示两类,如圄5所示(圄中与PMU、RTU相连模型为电力系统数据表示模型;与SmartMeter相连模型为个人用户数据表示模型)智能电网数据表示模型目前,电力系统数据描述已有的常用模型标准包括IEC60870协议组①、IEC61850协议组②、IEC61970协议组③以及正在制定的IEC61968协议组④。其中IEC60870协议组是较早(19901995年)制定的电力系统自动化协议组,其通信模型和数据模型适用于采用专用通信线路搭建的点对点通信网络,目前正在逐步被替换。 IEC61850协议组是描述变电站内通信网络和系统标准体系的协议组,于1999年发布。协议采用了面向对象的数据建模方法,实现了对数据的自我描述,传输的数据自己带有说明文件,使得数据传输时不需要再实现进行规约和转换,从而具备了面向服劳的特点,而IEC60870协议组下数据传输时需要收发双方事先对数据库进行规约IEC61970协议组及IEC61968协议组均针对电网调度管理系统,其中IEC61970协议组主要面向EMS(能量管理系统)。而IEC61968主要面向DMS(配电管理系统),上述两个协议组均采用了通用信息模型CIM.CIM模型也是采用面向对象的方法描述电网模型及其数据,可用UML图来表示电力系统组件间的继承、连接关系及资源属计算机学报性,同时CIM模型还定义了CIM/XML文件,使得CIM模型可通过XML进行传递,这样不同的应用系统就可以直接相互通信,因此CIM模型可用于电力系统的应用集成。同时,CIM还具有元数据描述管理的功能,可用于电网数据仓库的建立。采用CIM模型对电力系统及其数据进行建模是构建智能电网信息网的趋势,均提出了基于CIM模型的智能电网信息共享平台设计方案。

就以智能电表为单元的个人用户数据而言,已有数据模型有DLMS/COSEM模型,其对应的国际标准为IEC62056协议组①。DLMS对智能电表数据的读取、计费和负载控制进行了规约,COSEM涵盖了DLMS规约的传输与用户层规范。

5.2.2智能电网数据存储模型智能电网具有可靠性要求高和数据海量的特点,这要求智能电网数据的存储必须设置必要的冗余和备份机制;同时电网数据的存储模型必须满足快速查找和处理要求;而由于智能电网应用多样,不同应用实时性要求也不相同,由此智能电网的数据存储也可分为在线数据和实时数据两种模式。

目前主要有4种智能电网数据存储方案:第1类方案为多个数据集中器,单一控制处理节点加上利用关系数据库的集中存储。其中每个数据集中器负责从一定数量的量测设备中获取数据。目前我国电网系统中的广域控制模型与之类似;第2类方案与第1类方案类似,但将集中式存储拆分为分布式数据库存储。第3类方案取消了利用关系型数据库的存储模式,提出了基于XML的〈关键字,值〉模型,并且采用类似MapReduce的算法对数据库进行操作;第4类方案采用分布式文件系统与数据库结合的方式存储数据,即数据库中存储的不是原始的电网数据,而是数据的索引,原始数据以文件的形式存在于数据集中节点上,该方式类似于搜索引擎对网页的搜索。结合智能电网中家庭电力数据的存储和账单计算这一应用对上述4类方案的并发处理能力和处理时间进行了仿真并给出了结以及针对家庭月账单的计算时间。结果表明方案3的可扩展性较差而方案4的处理时间较长,方案1和方案2类似。

另一方面,由于智能电网数据应用类型数量不可预期,容易造成数据统一管理的困难。将智能电网数据抽象为历史模式、实时模式和未来模式进行建模,而不是按照应用类型对数据存储进行建模管理。其中实时数据管理主要针对实时数据分析的需求,利用内存数据库进行存储。历史模式主要针对历史数据的存储、查找,采用时序数据库进行存储。而未来模式主要用于存储未来的可能发生的设备的变化,例如加发电机等。在此基础上,上层应用可以按需获取和管理异构数据库,从而解决异构数据模型的管理问题。此外,还有探讨在量测系统AMI和数据管理系统DMS(DataManagementSystem)之间构建统一数据集成中间层MDI(MeterDataIntegration:59.从而使得AMI系统和DMS系统之间得到解耦,用于解决由于数据模型和通信协议的异构性造成数据存储和管理的困难。

数据存储模型选取的不同将导致查找、获取和数据处理模式的不同,同时也会引起系统响应时间的区别,如何为智能电网选取合适的存储模型,将是未来智能电网研究中的一个重要方向。

5.2.3基于云计算的智能电网数据存储从系统实现上来看,物联网系统的搭建依赖于云计算平台,云计算平台为物联网应用提供了计算和存储资源。作为物联网的一个典型实例,云计算技术与智能电网的结合是必然趋势。如提出了基于云模型的数据管理和处理模型,将智能电网数据分布式存储在电网的各个节点,然后以服劳的形式将数据提供出来供应用访问获取。云存储有助于解决智能电网数据存储的海量性和可靠性问题。

OpenPDC是目前已经按入运行的一个智能电网数据处理系统,其实现基于开源平台Hadoop.该系统应用对象为时间序列数据流,即数据源为经过GPS授时的数据流。应用背景即为智能电网中的WAMS系统,由于WAMS系统的采样频率为每秒30次,当WAMS系统的子单元PMU数量加时,会产生大量的数据。目前该项目管理了北美东部约120个PMU的数据信息,平均数据量约为每小时1.5GB.截至2009年我国仅在220kV电压等级以上电力系统部署的PMU单元个数已经达到1000以上,再考虑个人用户实时产生的数据,可以预见未来智能电网的数据量是非常巨大的。在这种背景下,集中的数据存储模式将对网络造成巨大的压力,采用分布式存储成为一种必然。同时由于电网稳定性的要求,数据本身存在冗余备份的需求。云计算平台的分布式文件系统可以为此提供解决方案,且有助于提高电网系统的安全性。

1期曹军威等:智能电网倍息系统体系结构研究如何将云存储应用于智能电网还存在不少问题尚待解决。首先,虽然已有提及未来电网的存储模型,但尚无较成熟的方案,数据采用数据库存储还是以文件形式存储仍有争议。其次,由于电网系统存在多样性的特点,不同量测系统的数据格式并不统一,例如不同厂家的RTU数据格式都不相同,如何构建统一数据模型的问题也需要解决。此外,失去时效性的大量数据需要迁移备份,并且这种迁移是频繁发生的,这种情况下保证存储系统的运行效率成为难点。此外某些应用对于电网数据的获取有时间限制,分布式文件系统的查找效率无法满足其需求。

总体来看,尽管云计算的分布式存储平台和并行处理模型适合未来智能电网分散性、可靠性、安全性和数据海量性的需求,但依然存在应用障碍。

5.3分析与决策系统智能电网按入实际运行后,面临的另一个巨大挑战就是海量数据的处理能力。由于智能电网既要满足个人络端用户与电网系统的交互需求,也要满足电网控制系统对电网稳定性的控制需求,未来智能电网中有两大类应用需要海量数据处理技术的支撑。第一类是智能电网稳定运行监控系统,它根据量测系统获取到的数据进行动态安全评估DSA(DynamicSecurityAssessment),保证电网运行稳定,以及电网系统出现故障后恢复系统。第二类是智能销售和消费系统,它通过实时电价自动平衡电能的供应和消耗,如微软开发的Google的PowerMeter系统②该类应用多与微网系统相结合,考虑新能源如风能、太阳能接入后分散发电资源的利用问题。此外,考虑智能电网数据的海量性,智能电网分析决策系统与云计算技术的结合是未来趋势,因此本文认为未来智能电网分析决策系统结构如圄6所示。

智能电网分析决策系统5 31智能电网分析决策需求对于第一类应用,第一是要解决电网稳定性的判定问题电力系统的稳定性分为静态稳定和暂态稳定两类,其中暂态稳定描述的是电网出现大扰动后的鲁棒性,比如出现短路故障、短线以及发电机突然摔负荷等等,如2003年美加大停电事故已有的电力系统暂态稳定评估方法(TSA)可以分为两大类,一类是基于数学模型的方法,包括时域仿真法,即通过建立电力系统各元件的微分方程,再通过数值方法求解各状态量的时间特性;基于Lyap稳定判据的能量函数法、扩展等面积法以及动态安全域法。另一类是基于数据本身的模式识别方法,包括神经网络、支持向量机、遗传算法等多种方法。其中第一类方法面临两个主要困难:一是实际电力系统规模很大,往往最后变成几干阶的微分方程求解,无法满足实时性要求;另一方面,由于电力负荷模型本身就是不可知的,现有分析方法往往采用估计和经验的方法给定负荷的参数,不精确,如何对电网负荷参数进行在线辨识也是未来智能电网亟需解决的问题。而第二类方法同样面临当系统规模较大时,数据集数量过大的问题,如何进行特征选取和压缩目前尚无统一的模式。另一方面,在获取到电网故障信息后,如何迅速重新配置电网结构使电网系统重归稳态是第二个需要解决的问题。已有的方法包括启发式算法、专家系统、数值计算、软件仿真及多级代理等。其中除多级代理之外的系统均基于集中式架构建立,当系统规模较大时会出现计算瓶颈。

通过智能电表获取到用户用电数据后,智能电网的另一项功能是对用户用电行为进行预测和建议,充分利用分布式能源发电能力,并通过电力使用时间的迁移降低峰值使用时间段电力系统压力,进而提高电力系统运行效率。其核心思想是利用实时电价调节用户行为。该类应用通常分三步实现:(1)根据用户数据构建行为模型并进行预测。

(2)中心处理单元获取用户数据进行全局优化(属于多目标优化问题),例如对于单个用户来说优化目标是最小费用,而对于电力系统来说优化目标是电力系统运行稳定性和效率。

(3)实时控制系统控制电器开关已有的实现方案多基于多级代理(agentbased),计算机学报每级代理进行出价(需要/发出的总电力及价格),再逐级汇总由最高级代理进行优化,如给出的实时定价算法就是这样一种方法。结合电冰箱的用电控制实例进行了说明。据,该类优化问题属于NP芫全问题,因此多采用启发式算法求解。此外,如不考虑用户向电网中送电的问题,则可以利用线性规划方法求解。

5 32基于云计算的智能电网数据处理由于电网系统规模大、节点多,特别是智能电表得到的数据需要实时规划和调度,这需要大量的计算资源进行分析处理,智能电网数据处理与云计算技术的结合成为必然。

已有研究工作探讨智能电网与云计算技术的结合,如将云计算的分布式数据存储模型和并行处理模型用于存储电网数据,对数据子集进行并行处理再汇总处理结果。上文提到的Hohm系统就是基于云计算平台的不足之处在于其处理算法要求数据子集之间互不相关,每个数据子集可以独立进行运算处理,智能电网中的某些应用符合这种运算模式,比如实时电价计算。但还存在一类应用,需要跨区域的数据分析才能给出结果,数据子集之间不能解耦,如调度、发电负荷平衡、电网应急报警。这种情况下简单的云计算模型并不能进行处理。电网的物理特性是系统本身的关联性较强(电力系统之间存在电气连接),也即意味着数据存在关联性,是否可以改进并行算法,降低传输和计算的资源消耗,是未来智能电网研究的一个方向,如前文提到的通过分析电网连接的耦合层度来降低数据传输量。此外,如何在松耦合系统模型下保证系统处理性能,满足处理时限要求,也是难点。

5.4控制与执行系统智能电网包括电能的发、输、变、配、用等5个环节以及分布式新能源的接入和使用,所以其控制系统在传统的厂站式控制系统上加入了额外的分布式频率、功率、电压、相位、负荷是电力系统的主要参数,电网系统频率下降、电压下降、发电机失效、过负荷都会造成电力系统事故甚至崩溃。传统的电力系统控制主要针对以上参数进行调控,具体包括稳定控制、电压及无功功率控制、频率及有功功率控制、配电网控制、柔性交流输电控制,在新能源大量引入后,分布式能源如何与传统电网结合是未来智智能控制执行系统能电网需要解决的重点问题,因为新能源接入往往会给电网带来新的安全稳定问题。在电压及无功功率控制方面,已有算法包括优化问题求解的梯度类算法、牛顿法、二次规划法、线性规划法以及模拟退火算法、遗传算法、蚁群算法及人工神经网络等多种方法。频率及功率控制方面,已有算法包括经典的IP控制、鲁棒控制、神经网络、遗传算法及线性规划法等。而配电网控制方面,已有算法包括整数规划法、分支定界法、混合整数法、人工智能和启发式算法以及基于多代理系统的方法。柔性交流输电控制主要基于静止无功补偿器ASVC、可控串联电容器补偿TCSC、可控移相器TCPS及综合潮流控制器UPFC.电力系统稳定控制和分布式能源发电控制的方法将在6.2节及6.3节进行详细论述,在此不做讨论。

从系统构架上来看,传统电网的控制模式多采用集中式的构架。所谓集中控制就是所有采集到的数据统一发送至数据中心进行集中处理并给出控制反馈,而分散策略指将大电网按区域划分,每个区域有自己的控制中心,控制中心之间通过共享数据实现对整个系统的控制。从系统性能上来看,集中式控制往往会对主节点产生过大的处理压力和带竞压力,同时也容易造成单点失效,所以未来电网的控制结构会逐步向分散结构过渡;另一方面,随着新能源的引入,未来电网将是许多分散的微电网的集合,分1期曹军威等:智能电网倍息系统体系结构研究布式控制的应用是一种必然。此外,分散式的控制模型下由于数据无需芫全在广域范围内传递,对于减少网络延时和保证网络稳定性也可能产生积极作用,如在信息网构架采用OPGW组网方法下研究了集中和分散控制策略下电网系统的延时和稳定性。从结果上看,分散控制的平均延时更小且方差更小,意味着网络的稳定性更好。

本节对实现智能电网的4个重要支撑平台进行了分析。为了实现智能电网对电力的稳定控制、能源的实时调配以及新能源的接入等目标,需要构建基于智能电网基础设施和支撑平台的应用体系。下面着重从发电侧、电网侧和用电侧这3方面对智能电网信息系统应用体系进行介绍。

6智能电网信息系统应用体系6.1发电侧应用由于传统化石能源的不可再生性及对环境造成的影响,绿色能源即新能源发电以及随之产生的微网系统正逐渐成为未来电网系统发展的趋势。

截至2008年,新能源占全球能源消耗的比例为19%,而且这一比例还在逐年上升。20042()()9年间,全球新能源容量的长速度在每年10%60%之间。列出了至2020年世界各国新能源发电占电网发电容量的预期百分比,其中丹麦、瑞士的部分地区预计在2030年前可用新能源发电芫全取代传统能源发电。

本节着重对新能源发电接入传统电网后的控制管理和能源调度问题进行分析,对于传统电网的发电控制问题不予涉及,因为该类问题已经在电力系统领域进行过多年的研究。

6.1.1新能源接入管理广义上的新能源包括可分派能源(cHspatchableenergy)和不可分派能源(nondispatchableenergy),其中水电站、生物能和地热能均属于可分派能源,而风能、太阳能和潮汐能均属于不可分派能源。划分的依据在于可分派能源的能源供应基本是可控的,而不可分派能源则相反,例如风力发电中风的速度和时间是不可控的。可分派能源的接入管理与化石能源发电系统无明显不同,而不可分派能源由于能源供应的波动性,接入电网后会对电网系统的稳定性出现影响。因此,不可分派能源的接入管理问题将是未来智能电网发电系统的研究重点。目前,不可分派能源发电主要以风力发电、光伏发电和燃料电池为主。据2009年的统计数据,风力发电的容量长比其余新能源发电系统容量之和还要多,全球风力发电装机总容量达到160GW,而光伏发电则是长速度最快的新能源发电系统。

光伏发电系统和风力发电系统的主要特点在于其能源供应的间歇性,因此会造成发电输出电压、频率的波动。而这种波动性在接入电网后会对电网系统的整体稳定性产生影响。这种发电电压和频率的波动性表现为两类问题:一类是正常发电期间由于能源供应波动造成的电能质量问题,如风速时大时小造成的电压不稳定;二是能源输入不稳定造成的能量输出波动问题,如风力发电中风机输出功率的波动,极端情况下风力过小或过大为保护风力发电机会停止发电,即停止输出。为解决第一类问题,电力系统领域已进行了数十年的研究,目前已有方法多基于电力电子器件的应用,通过在风力发电机和电网之间加入变流器及电容器组合等电力电子器件,以实现对电压抖动、频率抖动、无功补偿和有功输出等发电系统关键参数的控制。例如风力发电机自20世纪80年代起经过了四五代的改进,早期的风力电机速度不可控,风机输出仅通过一个无功补偿环节就加入到大电网中,因此风力的波动直接会输入到电网系统中,而目前的可控变速恒频风力电机已可较好实现对风机输出的电能质量控制,详细信息可。在电能质量控制方面已有技术包括机械开关电容MSCs、基于可控晶阐管的静止无功补偿SVCs以及静止同步补偿STATCOM在实现了在不可分派能源控制基础上,不可分配能源接入的稳定控制运行监控与上文论述传统电力系统控制模型的一致而对于不可分派能源发电间歇性造成的第2类问题,则更依赖于智能电网传感、量测、通信和数据处理环节的支撑圄8以风力发电为例描述了智能电网信息系统对风力发电接入的管理。

目前,解决风力发电系统可能出现的输出不稳定问题,主要有两条思路:一是通过预测风场所在地的风力输出信息,结合负载测的能源需求信息,通过与电能存储结合的混合新能源发电系统进行实时调度,以实现稳定的发电输出但第1种思路需要大量的分布式存储设备与风电系统配套建设,成本较高;二是通过对电网负荷的实时控制、平衡风力发电输出和负载功率需求之间的关系,在风机输出减少时减少负荷的使用,从而降低存储设备的规AirEnergyStorage)、超级电容及电动汽车储能系例如就是采用抽水储能与新能源发电系统结合的实例。系统不同的存储系统能量转化效率、建设成本和适用模式都不相同,详细内容可以

1. Introduction of Underground Oil Distillation Plant

Underground Oil Distillation Plant ,is the new technology which can refine the underground oil into base oil(which can be made into diesel and gasoline after processed by our catalyst) by vacuum pressure distillation system. The oil quality is better than the original normal pressure distillation technology, which show on purity ,transparence, lightness .this technology will do deodorization and destinke process to the raw material oil automatically by "dry type" vacuum pressure distillation method. With the vacuum distillation technology, the distillation temperature is considerably reduced, and the oil output will higher 5%-10% compared with original normal pressure distillation technology. It makes more profits to the enterprise virtually.

 

2. Raw material which can be used

    a. Waste oil .example: waste diesel, waste oil residue etc.

    b. tire/rubber oil

    c. plastic oil

    d. crude oil

e. waste engine oil

f. waste motor oil

g. waste lube oil

h. waste transformer oil

i. underground oil


3. Models of  underground oil distillation plant


 4. Installation: We will be in charge of arranging our engineer to go to your place to guide the installation and train your workers how to operate the underground oil distillation plant ,and buyer will be in charge of the food, accommodation and round air tickets.

 

 5.Underground Oil Distillation Plant Exporting Experience:

 America:

Brazil, Canada, Colombia, USA,

Middle East:

Dubai, Iran, Jordan, Saudi Arabia, Turkey

Europe:

Albania , Bosnia and Herzegovina

Asia:

Afghanistan, India, Malaysia, Pakistan, Philippines, South Korea, Vietnam, Myanmar

Africa:

Ghana, Mozambique, Zambia

 

Underground Oil Distillation Plant

Underground Oil Distillation Plant,Oil Distillation Equipment,Fuel Oil Refinery,Diesel Oil Recycling Distillation Plant

Shangqiu Sihai Energy Technology Co., Ltd , https://www.sihaienergy.com