科技通报
科技通報
과기통보
BULLETIN OF SCIENCE AND TECHNOLOGY
2015年
4期
239-241
,共3页
正态相关%高铁车厢%温度传感器%数据挖掘
正態相關%高鐵車廂%溫度傳感器%數據挖掘
정태상관%고철차상%온도전감기%수거알굴
normal correlation%high-speed rail compartment%temperature sensor%data mining
通过对高铁车厢车载温度传感数据状态模式的准确挖掘，实现对高铁车厢温度的精确控制，提高高铁运行性能。传统方法中对车载温度传感器的数据状态模式挖掘采用的是半导体气敏传感器测量法进行数据采集，并采用神经网络控制法实现数据状态模式挖掘，方法受限于温度数据的均衡控制无法准确把握，控制效果不好。提出一种基于正态相关的高铁车载温度传感数据状态模式挖掘方法，首先进行温度传感数据采集的硬件系统设计，在原始数据采集的基础上，提取温度传感数据的正态相关状态信息特征，采用PID控制器进行列车车厢的温度调节，温度变化率、积分时间和微分时间通过线性组合的方式进行控制，得到温度传感数据的自相关控制状态方程。数据跟踪器设计成二阶离散数据跟踪器，以满足实际需要。实现了温度传感数据的状态模式挖掘改进。仿真结果表明，算法温度控制和功率控制的稳定性和准确性优越传统算法，车载温度智能调节和控制性能提高。
通過對高鐵車廂車載溫度傳感數據狀態模式的準確挖掘，實現對高鐵車廂溫度的精確控製，提高高鐵運行性能。傳統方法中對車載溫度傳感器的數據狀態模式挖掘採用的是半導體氣敏傳感器測量法進行數據採集，併採用神經網絡控製法實現數據狀態模式挖掘，方法受限于溫度數據的均衡控製無法準確把握，控製效果不好。提齣一種基于正態相關的高鐵車載溫度傳感數據狀態模式挖掘方法，首先進行溫度傳感數據採集的硬件繫統設計，在原始數據採集的基礎上，提取溫度傳感數據的正態相關狀態信息特徵，採用PID控製器進行列車車廂的溫度調節，溫度變化率、積分時間和微分時間通過線性組閤的方式進行控製，得到溫度傳感數據的自相關控製狀態方程。數據跟蹤器設計成二階離散數據跟蹤器，以滿足實際需要。實現瞭溫度傳感數據的狀態模式挖掘改進。倣真結果錶明，算法溫度控製和功率控製的穩定性和準確性優越傳統算法，車載溫度智能調節和控製性能提高。
통과대고철차상차재온도전감수거상태모식적준학알굴，실현대고철차상온도적정학공제，제고고철운행성능。전통방법중대차재온도전감기적수거상태모식알굴채용적시반도체기민전감기측량법진행수거채집，병채용신경망락공제법실현수거상태모식알굴，방법수한우온도수거적균형공제무법준학파악，공제효과불호。제출일충기우정태상관적고철차재온도전감수거상태모식알굴방법，수선진행온도전감수거채집적경건계통설계，재원시수거채집적기출상，제취온도전감수거적정태상관상태신식특정，채용PID공제기진행열차차상적온도조절，온도변화솔、적분시간화미분시간통과선성조합적방식진행공제，득도온도전감수거적자상관공제상태방정。수거근종기설계성이계리산수거근종기，이만족실제수요。실현료온도전감수거적상태모식알굴개진。방진결과표명，산법온도공제화공솔공제적은정성화준학성우월전통산법，차재온도지능조절화공제성능제고。
Accurate mining on High-speed Rail carriage vehicle temperature sensing data state model is constructed for the precise control of the High-speed Rail compartment temperature, improve the High-speed Rail performance. The tradition?al method of mining is used in measurement of semiconductor gas sensors for data acquisition, and the neural network con?trol method is used to achieve the state of the data mining, balance control method is restricted, the temperature data cannot be accurately grasped, control effect is not good. A state mode mining algorithm of High-speed Rail on-board temperature sensing data is proposed based on normal correlation, temperature sensor data acquisition of original data collection is ob?tained, normal state information characteristic temperature sensor data is extracted to train the PID controller temperature control, temperature change rate, the integral time and differential time is controlled by a linear combination of the way, the autocorrelation control equation of state temperature sensing data is got. Data tracker is designed into two order discrete da?ta tracker, to meet the actual needs. The mode of temperature sensing data mining is improved. The simulation results show that the stability and accuracy is better than the traditional algorithm in temperature control algorithm and power control, in?telligent vehicle temperature regulation and control performance is improved.
