Application of comprehensive intelligent control technology in power grid planning

Guangdong electric power industry has entered the era of large power grid, high voltage and large units. How to rationally distribute the power grid is one of the important topics in Guangdong electric power industry. Power grid planning is a multi-objective system engineering which is constrained by many conditions and takes the total benefit of power grid as the ultimate goal. Therefore, comprehensive intelligent control technology should be adopted to study power grid planning. A successful example of this is the DC/ AC transmission network design expert system of Quebec Hydropower Company in Canada, which has the characteristics of target and expected benefit, interaction between domain experts and knowledge engineers. According to the current situation and development goal of Guangdong power grid, the power grid planning decision-making system can be decomposed into sub-problems, such as load forecasting, power grid planning, reactive power planning and stability analysis. Through the iteration of sub-problems, the optimal power grid planning decision-making system can be sought.

Guangdong electric power system includes 2 1 local power grid, and the existing maximum operating voltage level is 500 kV. A 500 kV ring network has been formed in the Pearl River Delta region, with 500 kV connected to Guangxi and 400 kV connected to Hong Kong and Macao, which are 65,438+065,438+00 kV respectively. In addition, Guangdong Power Grid also supplies power to Yizhang and Linwu counties in Hunan and Gannan area in Jiangxi.

Central Guangdong (Pearl River Delta) ground network is the core of Guangdong power grid and the largest load center in the province. The power grid is interconnected with Guangxi, Hong Kong and other power grids, which not only supplies power to the Pearl River Delta, but also undertakes the task of power exchange. Building a powerful 500 kV power grid in central Guangdong is of great significance to ensure the safe operation of Guangdong power grid, Hong Kong power grid and Macao power grid. At present, Guangdong 500 kV power grid has been extended to the west wing of Shantou, and the 500 kV power transmission and transformation project from Jiangmen to Maoming is under construction and is expected to be put into use before 2000.

Guangdong electric power industry has entered the era of large power grid, high voltage and large units. As the whole power grid becomes more and more complex, the previous artificial assumptions and local optimal planning methods will bring hidden dangers to the operation and development of the power grid, and the possibility of blindly using funds will increase. Combined with the development of current theory, we think that power grid planning is a multi-objective system engineering, and the ultimate goal is the total benefit of power grid, which is constrained by various conditions. For such a system, we think it is appropriate to study it based on cybernetics and combined with information theory, operational research and system engineering.

From the perspective of cybernetics, power grid is a typical dynamic large-scale system with huge dimensions, strong nonlinearity, time-varying and parameter uncertainty, and contains a large number of unmodeled dynamic parts. In addition, the power network is widely distributed, and most components have complex physical characteristics such as delay, hysteresis and saturation, so it is extremely difficult to realize effective decision control for such a system. On the other hand, due to the growing public dissatisfaction with new high-voltage lines, the increasing cost of the right to use lines, especially corridors, and the growing power network, people have put forward higher and higher requirements for decision-making control of power grids. It is precisely because of this characteristic of power grid that some advanced cybernetics ideas and technologies are continuously introduced into power grid. The following will explain the necessity and feasibility of introducing integrated intelligent control technology into power grid planning.

1 integrated intelligent control technology

1. 1 the concept of intelligent control

Up to now, there is no unified concept of intelligent control, and the document [1] is summarized as follows:

A) Professor Fu first put forward the concept of intelligent control. He summarized intelligent control as the combination of automatic control and artificial intelligence through the study of man-machine controller and robot. He believes that the bottom control should use the conventional basic controller, and the high-level intelligent decision-making should have anthropomorphic function.

2) On the basis of Fu's work, Saridis put forward an intelligent control theory system with ternary structure. He believes that only binary combination is not conducive to the effective and successful application of intelligent control, and it is necessary to introduce operational research to make it a ternary combination, and puts forward its theoretical framework of hierarchical intelligent control.

C) After studying the above theoretical structure, Professor Cai Zixing proposed a four-component price system in 1986 from the perspective of the integrity and purpose of the system, and summarized intelligent control as the intersection of control theory, artificial intelligence, operational research and system theory.

In a word, intelligent control is a combination of multidisciplinary knowledge. In addition to studying from the perspective of cybernetics, it can also be discussed and studied from the perspective of information theory, biology and social science.

1.2 integrated intelligent control technology

On the one hand, comprehensive intelligent control includes the combination of intelligent control and traditional methods, such as fuzzy variable structure control, adaptive fuzzy control, adaptive neural network control and neural network variable structure control. On the other hand, it contains a variety of intelligent control methods, such as expert fuzzy control, fuzzy neural network control, expert neural network control and so on.

2 foreign power grid planning expert system

So far, there are not many successful integrated intelligent control technology systems in power grid planning, among which Hydro-Quebec's "DC/ AC transmission network design expert system" is a better one.

In the late 1980s, with retirement and long-term non-use, a great deal of professional knowledge of power system planning and design gained by engineers during the rapid development of Canada's power grid in the 1960s and 1970s was gradually forgotten, which attracted the attention of Canada's power sector. Quebec Hydropower Company believes that expert system technology is a potential method, which can be used to express and preserve some specialized technologies and knowledge in the minds of human experts. They believe that in the field of power system planning and design, the loss of professional knowledge is very obvious, especially in the period of slow growth of power system. These professional knowledge come from various disciplines and play an important role in the decision-making process of multi-level power system design. Some selection decisions, such as power generation type, power plant location, transmission type (AC /DC), voltage level, number and model of transmission lines and number and model of compensation equipment, must be carefully weighed according to some standards, including reliability, stability, steady-state performance, cost and environmental conditions. On this basis, experts from Quebec Hydropower Company developed an expert system for preliminary design of transmission network, which has the following characteristics.

2. 1 objectives and expected benefits

The main purpose is to study the possibility of imitating human expert behavior with expert system in the preliminary design of AC /DC transmission network. Systematic determination and expression of knowledge required for qualified design, including symbols and digital data, as well as principles, rules, standards, compromise methods and mathematical models to guide design. Qualified design is based on cost, environmental conditions, stability, reliability and design sensitivity or robustness. ES prototype should also guide users to interact with knowledge base by completing the steps required for design, and provide explanations of corresponding reasoning paths after reaching each intermediate step. The main benefits expected are:

A) Expert knowledge can be retained and passed on to future engineers;

B) Knowledge can be expressed in more concrete forms, rather than some vague and unfounded judgments;

C) More consistent results will be obtained;

D) Compared with human experts, ES can check and compare more schemes and get more economical designs;

E) With the help of reasoning and explanation, ES can be used as a teaching and training tool for future experts;

F) As a "consulting" means or tool to evaluate and improve the existing design, expert system will be very helpful to experts;

G)ES will be the pioneer of expert system family for designing all kinds of power system equipment, and as a model, extract more general design methodology from it;

H)ES plays the role of collecting knowledge, which is usually scattered throughout the design organization.

2.2 Interaction between domain experts and knowledge engineers

Knowledge engineers should have experience in power system analysis and design and artificial intelligence (AI). Practice has proved that the mixture of these two kinds of knowledge is very effective for extracting and concentrating expert knowledge from domain experts. Expert knowledge comes from power system planning engineers who have many years of experience in planning, designing and debugging large-scale engineering projects.

2.3 Evaluation factors of design The candidate design must meet the following conditions:

A) minimum fault recovery characteristics of a)DC system;

B) Allowable radio and harmonic interference requirements;

C) Minimum stability standard after failure;

D) Limit of stable voltage and reactive power supply;

E) transient overvoltage limit after load rejection;

F) Minimum equipment redundancy required for reliability;

G) Must be insensitive to the change of input data (robustness);

H) Must meet certain maximum cost requirements;

I) must be applicable to the existing technology.

The "DC/ AC transmission network design expert system" of Quebec Hydropower Company has been successfully applied for nearly ten years and is constantly developing and improving. With the rapid development of fuzzy technology and artificial neural network, the application prospect of integrated intelligent control technology in power grid planning is more and more broad.

3 decomposition and coordination of power grid planning decision-making system

Power grid construction is a capital and technology-intensive project, and the economic service life of lines and equipment is as long as several decades, so whether the network structure is reasonable will have a long-term impact on the technical performance and economic benefits of the power grid. The loss of a planning mistake is hard to recover in a few years. With the continuous development of Guangdong power grid, how to rationally distribute the power grid has become one of the important topics for the development of the current power grid and even the whole power industry.

There are a lot of decisions to be determined in power grid planning, and these decisions are interactive in time and space. At present, due to various conditions, it is impossible to consider it with a unified model. It can only be decomposed into relatively simple subproblems, and then coordinated by iteration between subproblems. According to different problems, power grid planning can be divided into load forecasting, power grid planning, reactive power planning, stability analysis and short-circuit current analysis.

4 conclusion

Power grid undertakes the task of connecting power supply and users. In addition, in order to obtain the greatest power supply reliability and economy, it also undertakes the task of connecting with the power systems in neighboring areas. Due to the large investment demand and long service life of power grid equipment, the power system is strongly restricted by "past weight". Therefore, it is the goal of power grid planning to seek the best investment decision and ensure the long-term optimal development of the whole power system.

Combined with the discussion in this paper, it can be seen that the power grid, a typical dynamic large coefficient with huge dimensions, has the characteristics of strong nonlinearity, time-varying and parameter uncertainty, and contains a large number of unmodeled dynamic parts. The control effect we want to achieve is a multi-objective, rolling optimization dynamic non-quantitative index (engineering benefit of power grid), in which knowledge representation and processing account for a large proportion. In this way, it is necessary to use comprehensive intelligent control technology to effectively organize a lot of knowledge about power grid planning, optimize operation and get the optimal decision. At present, Guangdong Electric Power Industry Bureau and School of Electric Power, South China University of Technology have jointly carried out the theoretical research on "Expert Decision System for Power Grid Planning", and it is expected that an effective power grid planning decision system based on comprehensive intelligent control technology will be developed in 2000, which will play a positive role in promoting the power grid construction in Guangdong Province.

refer to

1 Huang Suan, Shao Huihe, Zhang. Intelligent control theory and method [J]. Control theory and application, 1994(4)

2. Wu Jingchang and Meng. Large power grid system technology. Second edition [M]. Beijing: China Electric Power Press, 1995.

3 Zhou Lerong. Research on object-oriented fuzzy database in power grid planning decision-making system: [dissertation] [D]. Guangzhou: School of Electric Power, South China University of Technology, 1998.

For reference only, please learn by yourself.

I hope it helps you.