Research On Multi-sensor System Bias Estimation In Complex Environment And Sensors Wake-up Strategy

In the complex systems,The ability of target detection and tracking is one of the key standards for judging the performance of antimissile and warning system in modern battlefield.However,system model mismatch caused by multiple complex uncertainties such as environmental interference,electromagnetic countermeasure,jamming deception,modeling bias,multi platforms bias and so on becomes the major constraint of system detection and tracking capability.Therefore,system bias modeling and the joint estimation of bias and target state has becomes the focus of current research,and on this basis,how to further efficiently achieve multi-sensor network collaborative detection and effective access to information has become the current challenge problem.Therefore,in this paper,the problems of stochastic dynamic system deviations,joint estimation of target states and unknown inputs,and cooperative wake-up strategy of nodes in multi-sensor networks are investigated in detail.The main contents are as follows:1.For the modeling mismatch problem,a system deviation model based on inequality statistic constraints is established,and a minimum upper bound filter based on convex optimization is proposed to realize the state vector adaptive estimation in the presence of inequality statistic constraints.An example of gyro rate sensor is used to verify the effectiveness of the algorithm.According to the characteristic of slowvarying of the bias drift in the output signal,the bias drift is treated as variance bounded unknown system disturbance.The influence from such bias drift to the upper-bounds of variances of estimation errors through introducing a free parameter,and then adaptively optimise this scale parameter for minimum upper-bound to obtain the gyro rate estimation in real time.This minimum upper bounded filter avoids the gyro drift modeling and meets the requirement of implications much easier.2.For the system mismatch with unknown statistical properties which can not be modeled as bounded variances,a generalized system bias(GSB)model is proposed,which is represented via a dynamic model driven by unknown inputs(UI).Based on GSB,a online joint estimation of system state and system bias on signal sensor is presented in two steps.First,the UI-free system bias model is derived.Then the linear minimum mean square filter is obtained via original system equation is converted into an equivalent UI-decouple system.The proposed algorithm can effectively remove the influence of the unknown input from the system,and obtain the optimal estimation of the state and sensor bias in the minimum variance sense.3.For the system mismatch with unknown statistical properties in the multi-sensor system,the algorithm for online joint estimation of system state and system bias on signal sensor is extended to the heterogeneous sensor network.First,the local system state and bias estimation is obtained through decouple the UI from the bias sensor measurements.Then based on the local estimation,the global estimation of system state and UI through network consensus.Finally,the global estimation can be used to refine the local sensor bias estimation.Then the joint estimation of system state,sensor bias and the UI are achieved.In this algorithm,by exchanging the sensors' local estimation with its neighboring sensor nodes,the network consensus strategy is used to improve the global estimation accuracy,and achieved the global consistent estimation.4.For the problem of information acquisition in large-scale distributed sensor networks,a biomechanical sensor-based wake-up strategy is proposed to realize the target-aware dynamic self-organization and target estimation.Inspired by the mechanism of immune Inspired by the immune mechanism after the organism infected by infectious disease,a artificial distributed infection-immunity model(DIIM)is presented based on five commonsense rules related to immunity characteristics.Different from the previous infection model,which only focus on the infection mechanism,the DIIM not only describing the infection mechanism,but also the immune responds and immune memory after infection.The resultant DIIM consists of six sub-processes reflecting the infection-immunity mechanism: occurrence probabilities of direct-infection(DI)and cross-infection(CI),immunity/immune-deficiency of DI and CI,pathogen amount of DI and CI,immune cells production,immune memory and pathogen accumulation under immunity state.DIIM can effectively improve the phenomenon of power-up sensor nodes can not be shut down after the target passing by.And is shown more more efficiency in energy consumption.