Font Size: a A A

A Study And Implementation On New Technologies Of Computer Decision-Making Support For Urban Power System Planning

Posted on:2004-06-28Degree:DoctorType:Dissertation
Country:ChinaCandidate:J XiaoFull Text:PDF
GTID:1102360155977540Subject:Power system and its automation
Abstract/Summary:PDF Full Text Request
With the rapid development of expansion and reform of urban power systems in China, urban power system planning has become an urgent task. Thus, systematical and profound research on theories in this field must be conducted, which can adequately consider the situations in China and meet practical requirements. Furthermore, intelligent computer decision-making support systems for urban power system planning are necessary to be developed, which are highly efficient, convenient and practical. Therefore, the dissertation has made research on some crucial problems in this area, with the application of new technologies such as artificial intelligence (AI), interval analysis and Multi-attribute Decision-making (MADM) theories, etc. Based on expert system, an intelligent process control system for urban power system planning is created. It integrates the powerful storage capability of relational database together with the flexible reasoning ability of rule-based reasoning (RBR) and cased-based reasoning (CBR). It can be used to manage complex planning processes and project status intelligently, result in obviously improvement on the efficiency of the planning. A pattern recognition based approach for automatic identification on connection modes of urban middle voltage (MV) distribution system is presented. It generates the model of network connection modes via topology analysis and creates a rule base for pattern recognition, then implements the network connection modes identification via characteristic matching method. Through the identification on typical connection modes of an urban MV distribution network and further analysis on the identification results, the structural characteristics of the system can be found. This method is very useful to planning, reconstruction and the standardization of connection modes in a distribution network. A new structure of multi-layer decomposition for spatial load forecasting (SLF) is built up, which divides forecasting units into gross load layer, data collection layer and simulation layer. This approach combines the advantages of both trending and simulation methods. In the proposed structure, the work in data collection and maintenance can be effectively reduced, under the condition of adequate spatial resolution in land decomposition and precision in SLF. An interval-based SLF model is also proposed based on the multi-layer decomposition structure, in which the consistent relationship among different layers is described in both spatial and temporal aspects. This model is capable of dealing with uncertainties in data collection and forecasting. It is fundamental work of building a unified interval-based model for the whole area of urban power system planning. An interval-based comprehensive method is addressed to provide engineers a powerful tool to accomplish spatial load calibration, which is extremely difficult and time-consuming work in practical SLF process. It is a method to estimate load distribution in case of incomplete information, which can be used not only in load forecasting, but also in the procedure of raw data processing. An interval analytic hierarchy process (IAHP) based multi-attribute decision-making (MADM) method in urban power system planning is presented. This approach significantly simplifies MADM problems in the planning, and combines the advantages of both human expertise and quantitative calculation. This method can successfully deal with uncertainties in MADM attributes and vagueness of experts'judgment, thus adapts to the needs of the flexible power system planning. A new comprehensive method of projects evaluation and decision-making is addressed, which is based on IAHP and interval 0/1 programming. It is capable of instructing project decision-making scientifically. This method will bring tremendous economic benefit to power network construction and has high practical value. Models and methods in IAHP are studied. With comparison and analysis of the results, accuracy and applicability of main methods on weights calculation in IAHP are evaluated. Further, the principle of applying of these methods to practical MADM in urban power system planning is concluded. At last, a probability theory based IAHP model is set up in order to describe the different fitness on values within the interval bounds. The theories and methods above are implemented in a practical computer decision-making support system for urban power system planning (CNP), developed by Tianjin University. The cases studied in this paper are all from practical planning.
Keywords/Search Tags:urban power system planning, intelligent technology, spatial load forecasting, multi-layer decomposition, interval algorithm, MADM, AHP
PDF Full Text Request
Related items