Search

For airborne electronic equipment, temperature is the most important factor affecting its performance, so it is very important to invest its dynamic temperature response process in a flight environment. By analyzing the heat exchange relationship between different devices in an electronic wing pod cabin, a temperature prediction method for electronic pod cabin based on the Random Vector Functional Link (RVFL) neural network is proposed in this paper. This method can complete a construction of prediction model using small amount of data. Hence it can realize a quick temperature response prediction only with the initial temperature values, and ensure relative high prediction accuracy.

Accidental environmental pollution threatens people's health and life badly, it becomes more and more significant to develop a method to identify an accidental source and predict its hazard accurately and quickly. An effective source identification method should be composed of contaminant detection, source location and strength estimation, meanwhile, contaminant detection is a key. The performance of source detection could determine the accuracy and rapidity of the subsequent source location and strength estimation. The method based on concentration threshold is a common detection method, which detects a contaminant source existing according to the observed concentration higher than the specified threshold. However, the accuracy of the real-time concentration sensor is worse together with observed noise. The detection method based on threshold could bring detection delay and cause false alarm with the observed noise. In order to solve these problems and improve the rapidity and accuracy of the contaminant detection, an accidental source detection method based on One-tailed U test is proposed in this paper. One-tailed U test is used to conduct the normal population parameter checking on accidental pollution. This novel method can overcome the influence of sensor noise, and improve detection performance and reduce the false alarm. In addition, it can ensure the reliability of subsequent source location and concentration prediction.

For maneuvering target tracking, general interacting multiple model (IMM) algorithms have a fixed structure. It has been overlooked that how the performance of these algorithms depends on the set of models used. The use of too many models is as bad as that of too few models. Therefore, a variable structure IMM (VSIMM) was presented and applied to ground target tracking. This algorithm eliminates the need for carrying all the possible models throughout the entire tracking period as in the standard IMM estimator, significantly improving performance and reducing computational load. But it is difficult to apply the VSIMM to other scenario (for example, aerial target) without auxiliary information such as map for ground target. This paper then presents a model set multiple hypotheses IMM algorithm (MS-MHIMM) which leads to a systematic treatment of model-set adaptation without using any additional auxiliary information. The new approach is illustrated in detail with an aerial complex maneuvering target tracking example.

A new control scheme based on a fuzzy-PI algorithm was presented to control the temperature control and check valve (TCCV) in a condensing heat exchanger (CHX) in manned module. There is a control threshold switching between PI and fuzzy control. When the absolute error is higher than the threshold fuzzy control is used, otherwise PI control is used. The algorithm uses instantaneous and longer-term changes in measured temperature to continually adjust the TCCV position and maintain the atmospheric temperature within specified limits. Simulation was conducted, and results showed that the proposed control scheme made it possible to improve temperature stability and control precision.

To identify the heat flux dynamically and predict the temperature more correctly, a particle filtering (PF) algorithm based on a double lumped thermal model is put forward. In the PF approach to the dynamic state estimation, one attempts to construct the posterior probability density function of the state based on all available information including the set of received measurements. Because the PDF embodies all available statistical information, it is a more effective method for the nonlinear estimation and prediction problem studied in this paper. Simulations were conducted. Results demonstrated the algorithm based on the double lumped thermal model could meet the precision of dynamical identification and real-time prediction for a satellite. The algorithm has a greater potential to apply autonomous control and self-adapting manage of satellite thermal control system in the future.