Design And Development Of Key Node Leakage Monitoring System For Heating Network

Due to forced assembly or soil settlement in the later stage of installation,the heating network is damaged,which leads to leakage accidents in the operation of the primary underground heating network.The research shows that the leakage of primary heating network will lead to the fluctuation of temperature,conductivity and internal pressure around the leakage point.There is a complex non-linear relationship between the leakage status and these physical quantities.It is impossible to evaluate the leakage status of pipeline comprehensively only depending on the change of a single physical quantity.In order to solve this problem,this paper designs and develops a high stability,low power consumption and high precision leak monitoring system for key nodes of heat network,which is distributed in the pipeline network site and can adapt to harsh environment.It can acquire real-time leak status information of pipeline network.Leakage monitoring system of key nodes in heat network is mainly composed of on-site monitoring unit and on-line monitoring system of primary heating network.Local monitoring unit completes field data acquisition and wireless data transmission,including pipeline state parameter acquisition module,microcontroller,communication module,power module and so on.High reliability sensors are used to acquire the ambient temperature and conductivity of the thermal pipe network and the operating pressure parameters of the pipe network on-line.RF radio frequency combined with GPRS wireless communication is used as the data transmission medium to share the acquired data to the data monitoring background.The data monitoring background is the upper computer software of the online monitoring system for the running state of primary heating pipeline network based on C#language.It displays and stores the data received from each monitoring unit,and predicts the data stored in the back-end SQL database using long-term and short-term memory neural network.Secondly,the leakage fault model of key nodes in multi-classification hot gate is established by the least square support vector machine(LSSVM)algorithm based on gravity search optimization.The state parameters reflecting the operation status of the heating network are used as training samples to train and classify and the prediction results are substituted into the classification model to evaluate the operation status of the network,guide the maintenance and prevent accidents.Thirdly,the first-order resistance-capacitance model is used to build a model for the supply lithium batteries of the leak detection terminal of the key nodes of the thermal network.Then,the battery circuit model parameters[k0,k1,k2,k3,k4,Cp,Rc,Rd]are taken as the model parameters of the extended Kalman filter algorithm.Recursive least squares algorithm is used to update the battery model parameters in real time.The residual charge state of the battery is analyzed and studied with the temperature,real-time current and voltage of the battery as input and SOC as output.When the battery power is too low,report it to the remote operating system in time,prompting the staff to replace the battery in time to prevent false negatives and false positives caused by low battery power.The key node leakage monitoring terminal designed and developed in this paper is installed and debugged in the well chamber environment of Huaneng a thermal power plant and has been put into use in batches.The field experiments show that the device has good application effect and meets the design requirements,which is the basis for online monitoring and accurate assessment of underground heating network pipe leakage.