NEURAL-NETWORK CLIMATIC PARAMETERS STRUCTURE FOR REFRIGERATED TRANSPORTATION
Many problems arise in the designing and controlling process of refrigerated transport because vehicles are viewed as immobile objects and the static climatic parameters are adopted. Considering that traditional mathematic method cannot meet the requirements of dynamic design climatic parameters, this paper analyzes several major parameters which intensely affect the climate, finds and develops the best BP（Back propagation）model by taking advantages of neural network such as the non-linear character and superior study ability, etc. Then, summer’s and winter’s outdoor dry-temperature as an example is calculated. After experimenting and proofreading work according to national standards, it shows that the error of the output in this model is very small as a whole in spite of some minor flaws in efficiency and partial accuracy. Therefore, the model can simulate the change of the outdoor climatic parameters better and is well worth applying.
Ruigui QU
School of Civil Engineering and Architecture, East China Jiaotong University,Nanchang 330013, China
国际会议
The 22nd International Congress of Refrigeration(第22届国际制冷大会)
北京
英文
2007-08-21（万方平台首次上网日期，不代表论文的发表时间）