| Water delivery using Pumping stations can effectively solve the problem of uneven spatial and temporal distribution of water resources, but the process of water diversion will consume a lot of energy, results in the high operating cost of pumping stations, for reducing the operating cost, it is significant to get better operating schemes from study the optimal operation of pumping stations. To optimize the operation of pumping stations is a complex issue, involving many factors, and inter-relations are very complicated, especially in multi-unit, multi-pumping stations running in parallel. It is difficult to calculate the problem and make decisions, so the reliability of problem models need to be improved, the efficiency and the stability of the calculating method need to be increased.The usual adjusting mode of operating conditions is variable angle adjustment and speed adjustment, and some studies have shown that, when taking into account the life and price of conversion device, the effect of optimal operation of pumping stations by variable speed adjustment is not obvious. So this paper only studies the optimal operation of pumping stations by variable angle adjustment, and the possible similarity during the process of pump blade adjusting is discussed. When calculating the problem of optimal operation of pumping stations, traditional deterministic algorithms are make sure to find the optimal solution, but the computational efficiency is very low; some researches for approximation algorithms only can be applied in the problem, focuses on the comparison between the optimal operation scheme and the design scheme, put less attention to how to improve the performance of algorithms according to characteristics of the problem, or avoid algorithm for trapped in local optimum and leave out the global optimal. Therefore, this dissertation achieves to study the optimal operation of pumping stations problem with the application of ant colony algorithm and pump blade adjusting similarity, for better strategies to solve the problem, and optimal schemes to guide the daily operation of pumping stations. The specific study may from the following aspects.(1) The possible similarity during the process of pump blade adjusting is discussed. We analyze the change of velocity triangles at the inlet and outlet of a pump impeller caused by blade adjusting, examines the flow behaviors in the pump passage, and derives pump blade adjusting formulae that reflect the variations of flow and pumping head with blade adjusting. Then we develop a mathematical model for calculation of the flow index and head index of these formulae based on the experimental data of pump performances, and formulate an objective function that is the least square sum of the formulae’s calculation errors relative to the test data. The calculation module of blade adjusting similarity was constructed, and an example was presented to verify its performance. With the pump performance parameters under design angle (0°), the blade adjusting similarity module can calculate parameters under any angle within the range of conventional variable angle, according to this, we are able to correct the performance errors which are caused of manual blade adjustment errors, and bring the angle adjusting performance closer to the real results.(2) Marco Dorigo and colleagues introduced the first ant colony optimization (ACO) algorithms in the early1990’s. ACO is one of the most recent techniques for approximate optimization, the inspiring source of ACO algorithms is real ant colonies. According to similarity between optimal operation of pumping stations problem and some other combination optimal scheduling problems which have application with ACO, an ACO model for optimal operation of pumping unit is proposed and the solution method by ants searching is presented by rationally setting the object function and constrained conditions. The heuristic information and the pheromone trail update method were improved by analysis characters of the model for better performance. A weighted directed graph was constructed and feasible solutions may be found by iteratively searching of artificial ants, and then the optimal solution can be obtained by applying the rule of state transition and the pheromone updating. The calculation module of optimal operation of single pump unit was constructed, to verify its performance, an example was presented and the result of single pump unit calculation module was compared with the result from dynamic programming or evolutionary solving method in commercial software under the same discrete condition, the result of single pump unit calculation module is better and the computing time is shorter.(3) Problem of optimal operation of multi-unit in pumping stations is complicated, and an efficient way must be found to solve it. Therefore, we developed an optimal multi-unit scheduling model which contains pump unit start-stop-once constraint, and proposed an ant colony collaborative hierarchical optimization algorithm to solve the model. First we discrete the water demand allocated to a single unit, take a single unit as a subsystem, using the single pump unit calculation module to find the objective of each subsystem; then according to the overall goal of the system, consider the relationship between the various subsystems, coordination modify the input and output of the subsystem, and ultimately to achieve the global optimization. For optimal schemes deciding, we proposed a projection pursuit evaluation method, selected the operating cost, the time of blade adjustment, the running time of the pump unit, operating efficiency and power consumption as the evaluation index, in order to analyze the problem of optimal operation of multi-unit in pumping stations comprehensively. The calculation module of optimal operation of multi-unit in pumping stations was constructed, to verify its performance, an example calculation was presented and the result of multi-unit calculation module was better, the final optimal scheme has less times of the blade adjusting, shorter operating time of pumps and lower electric charge, it is very suitable for daily operation.(4) We developed an optimal scheduling model of multi-pumping stations running in parallel, and proposed an ant colony collaborative hierarchical optimization algorithm to solve the model. First we discrete total water demand allocated to a single pumping station, take a single pumping station as a subsystem, using the multi-unit calculation module to find the objective of each subsystem; then according to the overall goal of the system, consider the relationship between the various subsystems, coordination modify the input and output of the subsystem, and ultimately to achieve the global optimization.The calculation module of optimal operation of multi-pumping stations running in parallel was constructed, consists of the computational method of the optimal scheduling model of multi-pumping stations running in parallel and the blade adjusting similarity calculation module. The multi-pumping stations calculation module can be applied to performance conversion under variable angle within the range of daily operation of pumping stations, and can calculate the problem of optimal operation of pumping stations with setting different adjustment step of the blade angle according to the actual regulation ability of pumping station, so that the optimal schemes are more appropriate to the actual daily operation of pumping stations. An example was presented to verify the performance of the multi-pumping stations calculation module, and the optimization effect of the module is obvious, which indicates that the module can provide a high application value to the field of optimal operation of pumping stations and related fields. |
PDF Full Text Request