Inexact Fuzzy Optimization Theory And Its Application In Energy Systems Planning

Energy systems planning is important in energy systems management, which can not only help accurately describe the behavior of the various activities in the systems, but also can quantify the possible effects of the planning schemes. It provides a scientific basis for decision-making management. However, energy systems usually consist of multiple subsystems, and there are complex interactions and relationships between each subsystem and among its internal components, which make energy systems present a series of complexity and uncertainty. Under this situation, a major challenge shows up in energy systems planning:the traditional deterministic programming models can hardly reflect those kinds of complexity and uncertainty. However, energy systems planning models based on uncertainty theory can comprehensively consider interactions among the factors of energy, economy and environment. At the same time, they could integrate the multiple complexity and uncertainty in energy systems into the model formulation and results computation, greatly improving the reliability and credibility of the planning schemes. Thus, researches on the application of uncertainty optimization theory in energy systems planning could provide practical technologies for energy systems planning and management, which is conducive to the efficient implementation of energy saving and emission reduction policies in China.In this thesis, fuzzy inexact energy systems planning models with the objectives being minimization of the total system cost, would be developed based on the comprehensive analysis of the complexity and uncertainty in energy systems by incorporating the improved fuzzy programming approaches into the interval and/or stochastic model frameworks, which could provide technical support for energy systems management. Specifically, the main research components include:(1) three fuzzy inexact programming models would be generated by integrating the lower-side attainment values based fuzzy programming into the interval optimization, chance-constraint programming and two-stage stochastic programming frameworks respectively to tackle the uncertainties expressed as interval numbers, fuzzy sets, probabilistic distributions, and the combination of them during energy systems planning. Moreover, compared with the conventional fuzzy programming, the lower-side attainment values based fuzzy programming would have an advantage in computation;(2) A variety of hybrid models including a feasibility based inexact fuzzy model for regional energy systems planning, a feasibility based inexact chance-constraint model for energy saving and emission reduction on power generation grid, and a feasibility based inexact multi-stage model for long-term energy and environmental systems planning would be proposed through the combination of interval or chance-constraint or multi-stage programming approaches with feasibility based fuzzy programming, which could quantify the violation during de-fuzzination compared with the lower-side attainment values based fuzzy programming;(3) Considering the energy planning problem in Jilin Province and the electric power system planning problem in Beijing as research objectives to demonstrate the applicability of the proposed models. A lower-side attainment values based fuzzy inexact model for energy systems planning in Jilin Province would be developed under the analysis of the characteristics of energy systems Jilin Province, which could help identify the optimal energy allocations with consideration of the complexity and uncertainty. According to the feature of the electric power system in Beijing, a feasibility based fuzzy inexact model would be proposed for the electric power system planning in Beijing, which could guide the capacity expansions of electric power generation technologies and the patterns of electricity generation in Beijing under the policies of energy saving and emission reduction.The obtained results indicated that the new developed optimization models would have advantages in uncertainty reflection and model solution compare with the traditional methods. Meanwhile, the shortcomings for each single traditional inexact programming could be remedied via the combination with other inexact optimization methods. The proposed models could optimize the allocation of different forms of energy, and assign energy activities under uncertainty, dynamics and interactivity. Moreover, these approaches could offer more flexible and robust support for energy systems planning in technologies and theory...