Research And Implementation Of Intelligent Selection Method For Non-Negative Pressure Water Supply Equipment

The popularity of high-rise buildings has promoted the development of secondary water-supply.Non-negative pressure water-supply measure has been widely taken because of its energy-saving and environmental protection.In the secondary water-supply system,the pump consumes too much electricity.The operation of the pump is closely related to the operating cost of the secondary water-supply system.At present,the determination of the flow rate in the selection of pumps in non-negative pressure water supply systems is often based on the “Building Water Supply and Drainage Design Specification”.The design flow rate is often higher than the actual water consumption because the national standard is universal.These inappropriate pumps lead to long-term and low-efficiency operation,resulting in a great waste of electrical energy.For one hand,excessive parameters in the pump selection are chosen manually by the designers,which is inefficient and wastes much manpower.The workload is large and the selection efficiency is low.For another hand,the problem has not been solved in comprehensive evaluation for the pump.The pump selection only depends on the comparison of energy consumption and efficiency,which is not comprehensive enough.Therefore,how to accurately calculate the design flow and optimize the pump selection method is an urgent problem to be solved in the design of the non-negative pressure water supply system.Thus,this paper has introduced the water demand forecasting mechanism,and established the comprehensive evaluation index of the pump.A multi-attribute decision-making method has been proposed to optimize the pump selection.The main research contents of this thesis are as follows:Firstly,the paper analyzes the collected water data,and determines the water concentration interval according to the analysis of the water law.The impact factors for building water flow are complex,the subjectivity of manual analysis is strong,and the nonlinear relationship between factors is not easy to find.A prediction model based on deep belief network is proposed.The factors related to building flow are used as input.The strong feature learning ability of the belief network automatically extracts the relevant characteristics of each influencing factor.The model was used to predict the building water flow,and compared with the BP neural network and the "building water supply and drainage design specifications".The experiment shows that the prediction result of the model is closer to the actual value and can be used to determine the pump selection flow.Secondly,on the basis of related theories of pump selection,the paper uses the rated flow and rated head to filter the pump,and calibrates the actual conditions after screening,which improves the efficiency of selection while ensuring the efficient operation of the pump.Aiming at the single method of water pump evaluation at present,this paper established a comprehensive evaluation system for non-negative pressure water supply equipment,and introduced multi-attribute decision-making technology to evaluate all primary selection schemes.The similarity degree of each scheme was calculated by TOPSIS method and then ranked according to this,and the best type selection result was obtained.This method is used to optimize the selection of a non-negative pressure water supply project in Changzhou.Compared with manual evaluation,this method considers it comprehensively and the selection results are more in line with the needs of users.Finally,based on the above research,the paper designs and implements a web-based intelligent selection prototype system for non-negative pressure water supply.Tests have shown that this system greatly improves the intelligent level of pump selection.