Study On Optimization Models And Key Technologies For The Whole Process Of Energy-saving Generation Dispatch

In order to improve energy efficiency of the power industry, save energy and reduceenvironmental pollution, Jiangsu, Henan, Guangdong, Guizhou and Sichuan were selected aspilots to enforce the energy-saving generation dispatch in2007. As the nationwideenergy-saving generation dispatch is carried out progressively, some positive effects onenergy saving and emission reduction have been achieved in the power industry, but there arestill some obvious shortages. Currently the energy-saving generation dispatch is focused onthe energy-saving generation scheduling algorithm based on the merit order of energyconsumption of each generation unit, i.e. optimizing the short-term generation schedule ofeach unit within the operational condition of the existing power system. In fact, the mediumand long-term dispatch shall be paid more attention in the energy-saving dispatch so as toachieve more significant energy-saving effect. In making the current schedule, energy savingis regarded as the first objective; in case the energy consumption is of the same level, thesecond-round rank is based on the emission level. In the real-time scheduling, the satisfactionof CPS standard is generally paid attention, while the requirement of energy-saving is ignored.These are the issues to be further studied in the current energy-saving generation dispatch.Meanwhile, currently most of the generation rank of coal-fired units is based on the designcoal consumption, and can not reflect the consumption level of the whole system actually,thus how to adopt actually measured data of coal consumption to implement generationranking should be studied.Therefore, this thesis proceeds deep research into energy-saving generation dispatch, inallusion to the shortages of current energy-saving generation dispatch model, analyzes thecurrent energy-saving generation dispatch model and method, does research concerning keytechnologies of load predicting methods and coal consumption online computation andproposes an optimized model for the whole process of energy-saving generation dispatch.Main research contributions are as follows:1. By analyzing the shortages of the current energy-saving generation dispatch model,research a new idea of the whole process of energy-saving generation dispatch covering year,month, day, real-time is presented. Using the curve characteristic of the load-coalconsumption rate of the coal-fired unit descending type, research and establish optimizedstarting up mode in the range of monthly schedule. By reasonably arranging units to beoff-the-line to improve load rate of starting up units, thus improve new type of energy-saving generation dispatch model. Design the time scale of year, month, day, real-time, etc. forpreparing procedure of energy-saving generation dispatch schedule and research coordinationand link up of all levels of generation schedule.2. Carry out research into load predict method for environment of energy-saving andemission reduction, compare and analyze model and algorithm of existing load predict,develop the load predicting method based on quantized meteorological condition influenceanalysis, establish practical sensitiveness quantized analysis method for the influence of airtemperature on load, research and propose to identify and amend method for electric loadabnormal data based on density estimation.3. The adoption of designed coal consumption to implement energy-saving generationdispatch ranking generates issue on fairness and effectiveness. The research adopts coalconsumption online computing model of anti-balance type of coal-fired unit, after computingthe actually measured coal consumption of coal-fired unit, put forward the method ofhandling discrete coal consumption data into the only coal consumption data under differentload of unit, and get the unit load-coal consumption rate curve and load-coal consumption ratecurve. After getting the coal consumption value and relevant index of unit, carry on researchinto three-level consumption difference index system of coal-fired unit, innovatively putforward the method of adopting high-dimension automatic optimizing method to confirm thedesired value of unit operating parameter. According to the coal consumption actuallymeasured data obtained, put forward the practical method for refined coal-fired unit ranking.4. Research into the monthly unit combination and daily generation schedule model ofthe whole process of energy-saving generation dispatch. Put forward emphasis on the monthlyunit combination model of unit shutdown, target function takes unit characteristic functioncombined by energy-saving and emission reduction characteristics into consideration,comprehensively consider the constraint condition of unit’s minimum shutdown time,maximum shutdown times, starting up cost, etc., optimize the monthly shutdown state ofcoal-fired unit. According to the requirements of energy-saving first, give consideration toenvironmental protection for energy-saving generation dispatch, establish two stages ofgeneration dispatch decision model based on layering ordination, the objective of the firststage is energy-saving, the second stage considers environmental protection. The effectivenessof the new energy-saving generation dispatch mode is verified through the computationalexample.5. Carry on research into AGC general directive dynamic optimization theory,innovatively put forward multiple target mathematical model of controlling target function of minimal power deviation and comprehensive energy consumption index and consideringvarious operational constraint, and adopt Q algorithm based on reinforced study to solve. Theoptimized algorithm which is based on Q-study can perform dynamic dispatch of AGCgeneral adjusting directive from dispatching center, proceed optimized dispatch to severalunits of different energy consumption, can make contribute total sum of each generator satisfyCPS standard well, and achieve the objective of energy saving and emission reduction byadding energy-saving generation dispatch index to reward function.Finally, the major research work carried out in this thesis is summarized, and prospectsfor future research briefly outlined.