Study On Optimization Identification Of Resistance Coefficient Of Metering Heating System

The flow rate of the heating network in the metering heating system changes with the independent adjustment of the user,the change of the hydraulic condition of the system is more complex,and the hydraulic balance is not easy to realize.Mastering the real-time resistance distribution of the metering heating system is helpful to study the dynamic characteristics of the metering heating system.The optimization identification method of pipe network resistance coefficient is an effective means to obtain the realtime resistance distribution of the metering heating system.The optimization identification problem of the resistance coefficient of the pipeline network is a multiobjective optimization problem,which puts forward higher requirements for the optimization algorithm.In this paper,the optimization identification method of resistance coefficient of the metering heating system and the optimization algorithm applied to resistance identification are studied.Firstly,based on the technical background that the metering heating system can transmit data remotely in real time,the resistance identification model based on flow measurement points is established.According to the variation characteristics of the resistance coefficient of each pipe section of the metering heating system,the pipe network is divided into two categories: outdoor pipe section and "variable resistance end pipe section".Then,an optimization identification scheme for the resistance coefficient of the metering heating system is designed.In the scheme,the observation data of multiple working conditions are preferentially used to identify the resistance coefficient of the outdoor pipe section,and then the resistance coefficient of the "variable resistance end pipe section" under each working condition is calculated.The reliability of identification results is ensured by using data of multiple working conditions.Second,the Differential Evolution algorithm is used to solve the resistance identification model based on flow measurement points,and the performance of the Genetic Algorithm and the Differential Evolution algorithm in dealing with the resistance coefficient optimization identification problem is compared.The results show that the distribution of flow rate error obtained by the Differential Evolution algorithm is more centralized and more stable than that of the Genetic Algorithm.According to the characteristic that the identification results of each outdoor pipe section identified by the Differential Evolution algorithm approximately obey normal distribution,a method of calculating the resistance coefficient of outdoor pipe section under a certain working condition by using multiple identification results of the Differential Evolution algorithm is proposed,and this method is combined with the optimization identification scheme of resistance coefficient of the metering heating system.The dual control of identification times and the number of working conditions used in identification is realized.Finally,the multi-objective optimization algorithm is applied to the optimization identification of pipe network resistance coefficient.A multi-objective optimization algorithm named GR-MODE is introduced,which is suitable for the optimization identification of pipe network resistance coefficient.The method based on fuzzy mathematics is used to select the compromise solution from the approximate optimal solution set as the most satisfactory solution.Through the comparison test of two actual cases with different number of pipe segments,the conclusion is drawn that GR-MODE is suitable for identifying the resistance coefficient of smaller scale heating pipe network.