Research On The Application Of Rbf Network Based On Particle Swarm Optimization In The Energy Consumption Monitoring And Management Platform Of Building Economical Campus

At present, most of the existing building has not installed energy consumption refined measuring equipment and partial energy situation is indefinite, the supervisors are shot of scientific evidence for estimation when they make energy-conservation measures and improve management mode. In order to solve this problem, this thesis take the project that "monitoring and management platform of buildings’ energy conservation for an economical campus of GuangXi University" as the background, carry out the research and design for building energy consumption online real-time monitoring system. According to the requirements of "Technical guide of institutions of higher learning campus building energy-saving supervision system’s construction" and the research results about the existing building in Guangxi University, research the system’s construction principle and construction target, design the overall framework of the system, networking mode, data acquisition and transmission mode, sum up the experience in building systems. Realize the monitoring to the energy consumption data of49building public buildings in Guangxi university.In the light of the problems of platform’s original data processing and analysis methods are over simple, parameters are hard to set up, results of the analysis are not practical, through research the energy consumption data processing and analysis method based on radial basis function neural network, and using particle swarm optimization algorithm to optimize the RBF network’s hidden node width and output weights, this thesis design a energy consumption data processing and analysis scheme that based on PSO-RBF algorithm. Select the college of foreign languages of Guangxi University as office building’s example, analyze and calculate the complex correlation between energy consumption data and its influence factors, determine this three volume:daily maximum temperature; working time; whether working day as its influence factors of energy consumption, design RBF network’s structure and PSO-RBF algorithm’s structure, construct the RBF energy consumption data analysis network trained by orthogonal least squares algorithm and the RBF energy consumption data analysis network trained by particle swarm optimization algorithm. By comparing the two group networks’analysis results, proving that the RBF network trained by particle swarm optimization algorithm has better performance, and is suitable for process and analysis building energy consumption, can help to improve the defects of platform’s original data processing and analysis methods. Take the original daily consumption alarm function of platform for an example, calculate the referenced threshold value of daily consumption alarm method of28office buildings by PSO-RBF algorithm, and set the threshold value of daily consumption alarm of college of foreign languages according to the results of analysis, make the software platform to reasonably play energy consumption alarm function, improve the data processing and analysis function of platform.