Swarm Intelligent Optimization Of The Load Distribution Of The Parallel Devices In Chilled Water Station Based On The I~2B Network Platform

A chilled water station is the core part of the air conditioning system and usually be composed of many large electromechanical devices such as pumps?chillers and cooling towers.It is an important part of the central air-conditioning system and also a large consumer of energy.As energy resources being increasingly scarce and the development of relevant technology,the focus of operation and maintenance of air conditioning chilled water stations has been shifted from simply ensuring the requirements of terminals to efficiency,energy-saving,safe and stable operation.However,the current control systems commonly used in chilled water stations built in centralized structure are confronting with problems such as poor practicability,poor universality,low construction efficiency,and difficult to be upgraded or reformed.In order to deal with this problem a corresponding "Insect Intelligent Building(I2B)" network platform has been proposed.Under the structure of this network platform,each device such as pumps,chillers,and cooling towers in the chilled water stations will be updated into an intelligent device possessing standardized and unified interface by embedding a Computing Process Node(CPN)inside.The original configuration process will be replaced by plug and play.Each intelligent device is an independent and equal individual,and the whole operating control process of the chilled water station will be handled by means of self-organization and communication between adjacent intelligent devices.This platform can effectively solve the problems existed in the centralized structure mentioned above.Since the operation mode of the algorithms on this platform has much difference with that on the traditional centralized structure.Existing studies on relevant optimization algorithms cannot be applied directly.Thus,it is of urgency to study and handle the optimal control of the electromechanical devices in the chilled water station under the I2B network platform.The optimal control problems of the chilled water stations can be abstracted in two kinds,namely the value set problem and the load distribution problem.Among them,the load distribution optimization problem is a kind of typical combinatorial optimization problem.This problem is existed in devices working in parallel mode which is commonly used in the chilled water stations.The optimization objective is to meet the demands of total amount of cooling or water supply with the minimum total energy consumption.The solution of the problem should be conducted by determining the allocation of load for each device.Based on relevant research progress,this article will focus on this problem by introducing the theory of fuzzy modeling,subtractive clustering,finite state machine and swarm intelligence.The contents can be summarized as followed:(1)Modeling methods are the basement of study the load distribution optimization problem.In order to solve the problems of the existed modeling methods of the devices in the chilled water stations,such as complicated and diversified forms,poor universality and difficult to be updated online,the fuzzy modeling method is introduced under the I2B network platform.The automatic extraction method of the fuzzy rules by utilizing the subtractive clustering is proposed based on which the online fuzzy modeling method of based on subtractive clustering is then proposed.The characteristics of the modeling method are simulated and verified,and then compared with the recursive least square method.Results have shown that this method can dynamicly update and extend the fuzzy model and possessing the advantages such as stable,simple in implementation,strong generalization ability.(2)The unified mathematical description of the supply quantity distribution optimization of three kinds of typical parallel devices is analyzed and summarized,including water pumps,chillers and cooling towers.The swarm intelligence optimization method is introduced and the unified solution paradigm based on the I2B network platform for the load distribution optimization is given.The distributed particle swarm intelligence optimization algorithm(FDPSO)and artificial fish intelligent optimization algorithm(FDAFSA)are proposed based on the I2B distributed unified solution paradigm.(3)Due to the characteristic of I2B network platform,the existing similar environment cannot meet the requirements of simulation validation of the above optimization algorithm.Therefore,the software virtual CPN node(VCPN)needs to be customized according to the basic function of CPN node.The proposed FDPSO and FDAFSA algoritlhm for the I2B network platform in the previous chapter have been validated on the simulation platform with VCPNs.(4)Based on the I2B virtual simulation platform,FDPSO and FDAFSA are applied to the optimization of the distribution of cooling supply among the parallel chillers.The results show that both FDPSO and FDAFSA are obviously more energy saving than the conventional operation strategies and that under the I2B architecture,FDAFSA has better stability and faster convergence speed than FDPSO.(5)FDPSO and FDAFSA are applied to the optimization of the distribution of water supply among the parallel pumps through simulation on the I2B virtual simulation platform.The results show that both FDPSO and FDAFSA are obviously more energy saving than the conventional operation strategies and that under the I2B architecture,FDAFSA has better stability and faster convergence speed than FDPSO.