In order to improve the performance of maneuvering target tracking, a RA-Jerk-EKF method is proposed by integrating the target radial acceleration into the measurement matrix, thus to improve the tracking accuracy by increasing the dimensions of measurement matrix.In the proposed method, the radial acceleration is derived based on Fractional Fourier Transformation (FRFT) in signal processing phase, and is then brought into the measurement vector through coordinate transformation.In data processing phase, a method of Extended Kalman Filter (EKF) based on Jerk model is adopted to resolve the problem of the non-linearity of the measurement equation in filtering.Simulation is made to compare the proposed method with the traditional Jerk model based EKF without acceleration measurement.The results show that the proposed algorithm has better performance on tracking accuracy, and estimation accuracies of position, acceleration and velocity are improved at the same time.

The integrated guidance and control design of ground missiles with angular constraints can effectively improve the guidance precision and destructive effect of missiles, which has become a hot topic in recent researches.Firstly, the integrated model of small air-to-ground missiles in pitch plane was established.It was transformed into a linear system containing angular constrain in order to achieve the angular constraint.Then, the linear quadratic optimal control method was used to design the integrated optimal control law of missiles with angle constraint.Finally, simulation was made to compare the designed control law with the sliding mode variable structure control law.The simulation results show that the optimal control law can not only ensure the accuracy of the missile guidance with angular constraint, but also can guarantee smooth trajectory and shorten the time of guidance, which shows good performance for missiles.

A new method for estimating the attitude of spinning missile was proposed.Firstly, according to the nature of spinning missile, a special sun sensor is designed to measure the solar vector instead of the commonly used accelerometer and magnetometer in MEMS sensor to correct the gyro bias.The solar vector measuring device comprises a transparent shell, photocell array, real-time clock and controller.When the sun light illuminates the photocell array through the transparent shell, the photocell array can convert the solar energy into the electric potential signal.According to the size of the solar energy and the angle of the photocell, combined with the accurate time provided by the real-time clock, the solar vector of this moment is obtained through main controller processing.Then, the Kalman filer state equation and observation equation established by using the quaternion differential equation is adopted to estimate the spinning missile attitude.The simulation results show that this method can obtain the higher accuracy attitude angle of spinning missile.

Aerial target intention recognition is an important part of the battlefield situation assessment, which is directly related to the operational decision-making of the commanders. However, the target's information measured in a complicated combat field is from multiple sources and involves many uncertain factors. In this paper, an aerial target intention recognition approach is proposed based on Belief-Rule-Base (BRB) and Evidential Reasoning (ER). Firstly, a new target intention recognition model based on BRB-ER for multi-source information fusion is presented. Then, a multi-parameter optimization model is established to optimize the initial parameters for improving the recognition precision. Finally, a case study is examined to validate the efficiency of the proposed approach. The result shows that it can recognize the aerial targets'intentions precisely.

In traditional quadrotor PID controller, parameter tuning is difficult and it is also difficult to achieve optimum control effect.To solve the problems, we constructed a quadrotor PID Neural Network (PIDNN) controller, which integrated the advantages of the traditional PID controller of clear engineering meaning and simple parameter tuning, with the advantages of Neural Network (NN) of nonlinear mapping and self-learning capability.The nonlinear mapping and self-learning capabilities of NN were used to optimize the control effect of traditional PID controller.By constructing the PID controller, the initial values of number of neural network layers, modes and connection weights were determined.At the same time, we designed a kind of adaptive flight control algorithm of PIDNN, using adaptive adjustment of proportional neuron weighting coefficient to increase the response speed of the system.The rationality and validity of the algorithm were verified by using a nonlinear full numerical simulation.

Feature detectors are receiving increasing attention from computer vision research community, which have been widely utilized in a large number of applications, such as wide baseline matching, object recognition and categorization, image retrieval, visual search, robot localization and data mining.To discuss research challenges of investigation and directions for further research, analysis is made to the robustness and speed of current widely used feature detectors.Various feature detectors, including Scale Invariant Feature Transform (SIFT), Speeded Up Robust Features (SURF), Binary Robust Invariant Scalable Keypoints (BRISK), ORB, KAZE, and Accelerated-KAZE are reviewed.Also, the algorithms affecting robustness and speed of feature detectors are investigated based on steps of feature detection.The Mikolajczyk 05 testing image sequences are used to determine and analyze repeatability and time cost.The experimental results show that fast nonlinear scale space, Features from Accelerated Segment Test (FAST) and long-distance pairs iteration have a more comprehensive performance.

To solve the problems of the traditional Sliding Mode Control (SMC) and traditional sliding mode reaching law, such as slow convergence speed, long convergence time and severe chattering in the course control of Unmanned Surface Vehicle (USV), a new Terminal Sliding Mode Control (TSMC) based on new power reaching law is proposed.Firstly, the non-linear function is introduced in the design of sliding hyper-plane based on the idea of TSMC, and a Terminal sliding mode surface is constructed.Under the assumption of improving the dynamic response by the TSMC, a novel power reaching law based on traditional power reaching law is introduced.Then, a new Terminal sliding mode controller is designed based on new reaching law and the TSMC theories.Finally, the Lyapunov stability is proved for the designed control system.The simulation results indicate that:compared with the traditional power reaching law and the exponential reaching law, the designed controller has faster convergence speed and better movement quality.

For target tracking carried out by Unmanned Aerial Vehicles (UAVs), CamShift algorithm is easily influenced by color similar background interference and sheltering interference, thus we made improvement to the algorithm.Firstly, considering that the tracking template of CamShift has little information and is easily influenced by color similar background, we proposed an improved CamShift algorithm based on 2D histogram template with H and LBP, by which the robustness to color similar background was improved, and the valid frame rate was increased by about 21%.To the problem of easily influenced by sheltering interference in target tracking, the Kalman prediction algorithm was introduced into CamShift algorithm, which could improve the robustness and tracking efficiency under sheltering interference, with the valid frame rate increased by about 25% and the average iterative time of per frame decreased by 36%.

When the warship is in the shore-based calibration process, the position of the laser rangefinder laid on the deck for targeting may not be accurately obtained due to the deformation of the deck, and position errors existed may seriously affect targeting accuracy.To solve this problem, a hybrid location algorithm is proposed.Firstly, the linearization is made to the nonlinear equation, and the least square solution is obtained.Then considering the position error together with the ranging error of the range finders, the suboptimal solution of the target position is obtained by using the weighted least square method.The suboptimal solution is taken as the initial value of Taylor series expansion method for iteration, and the final location can be obtained.The simulation result showed that:Compared with Constrained Total Least Squares(CTLS)algorithm and CHAN algorithm, the proposed algorithm is more reliable, and has an accuracy relatively close to CRLB even under high noise condition.

Considering that the sample fly-wing UAV has the characteristics of unstable course and multiple actuators, we established the converse model of the UAV based on neural network, which could transform the controller outputs into actuator offsets.Then, we analyzed the principle of snake maneuver from the aspects of rolling and yaw angles, and designed a snake maneuver control law based on Robust Servomechanism Linear Quadratic Regulator (RSLQR) control and Proportional Integral (PI) control, which could ensure the rapid response of roll and track angles while getting rid of the sideslip angle rapidly.Simulation results shows that RSLQR-plus-PI control can resist the disturbance, maintain the properties of the controller, and guarantee the robustness and stability of fly-wing UAV more effectively.

By taking advantages of the technical feature of the terminal guided weapon with search ability, area shooting method can search and attack outlined target effectively.The current researches analyze the shooting process and kill probability with the statistical simulation method, mainly concentrating on normal distributed target rather than the outlined target with larger errors.Area shooting can be subdivided into two basic issues:weapon amount calculation and aiming point configuration.To the outlined target with uniform distribution, the optimal weapon dispersion density function is deduced by constructing kill probability model, and a predictor-corrector method for calculating weapon amount is established, whose corresponding kill probability can reach the predetermined requirement.By setting the kill probability upper bound as index function, a solution for aiming point configuration is proposed.According to the technical details of contemporary weaponry, the numerical simulation is conducted to verify the validity and timeliness.Comparison with the existing method demonstrates the advantages of our method both in the extent of optimization and in the effectuation of satisfying predetermined kill probability.

In order to solve the extended target tracking algorithm in the case of inaccurate measurement, an ET-CBMeMBer filter algorithm based on box particle filter is proposed.Based on the stochastic set theory, the state set and the set of observations of the extended target are first randomized.Then, based on the interval analysis technique, the multi-extended target pseudo-likelihood function and the potential balance Bernoulli multi-extended target status updating function are deduced out.A fuzzy ART interval measurement set partitioning method suitable for interval measurement is proposed.Then, continuous and steady target tracking is implemented on the basis of the setting of measurement sets.Finally, the simulation experiment is carried out, and the results show the effectiveness of the proposed algorithm.

For the infrared early warning system, the multi-target tracking algorithm usually has long starting time and poor real-time performance due to long measurement cycle and large quantity of targets. In this paper, a new algorithm based on random finite set is proposed. Considering that for the standard multi-target tracking algorithm based on the standard random set, it is necessary to give the prior birth probability density of the new target and the track can't be labeled, we make improvement to the algorithm. Sector processing method, uniform new target density model, and track batch management are used. Simulation results show that the improved algorithm can meet the demand of infrared warning system.

A method suitable for engineering application is presented to determine the minimum sample size of missile-borne electromechanical products in storage environmental test.Firstly, PAVA, a method for processing “reversal” data based on isotonic regression, is used for data pre-processing.Then, the lifetime distribution model is established by using the storage reliability evaluation method combining minimum chi-square estimation with goodness-of-fit test.Finally, the relative deviation of estimator is used as the precision index, the minimum sample size under given precision requirement and confidence level is calculated out.A formula indicating the relation among sample size, precision index, confidence level and distribution parameters is derived.The result shows that the minimum sample size is fit for engineering application.

The synchronous mode of bistatic radar system is studied, and a method is proposed to parameterize the transmitter synchronization signal by using the GPS second pulse time information.The receiver can receive the echo signal by the reconstructed synchronization signal with the parameters.This method avoids obtaining information from the direct wave, and can improve the sharing of the transmitter, thus any one of the receivers can share the same transmitter.Through experiments to navigation radar JMA2254, radar parameter reconstruction with no direct wave is realized,which achieve the desired experimental results.

Aiming at the shortcomings of the static two-dimensional Johnson criteria of relying on manual interpretation and being highly subjective, we proposed a multi-dimensional distinguishable criterion, which can be calculated via computer online.The main purpose is to study the quantitative impacts of imaging angle, Distance Resolution (DR), Signal to Noise Ratio (SNR) and Line Pair (LP) on the probabilities of detection, recognition and identification in laser imaging.The data query form of total space style is given.Three recognition and identification algorithms that have universal nature are given, and the criteria results are visualized via highly realistic surface expression.Finally, a set of criteria conclusions that have high value of engineering reference are presented.

Field Programmable Gate Arrays (FPGAs) based on Static Random Access Memory (SRAM) are extremely sensitive to Single Event Upset (SEU) induced by radiation particles.In order to evaluate the dependability of the user designed FPGA circuit, a positioning fault injection methodology based on locating the configuration memory position is proposed for simulating SEU.The fault injection methodology is described from two aspects of configuration bits and user bits respectively.Fault injection tests are made on LUT and BRAM to prove the validity of the methodology.In this paper, the fault injection method, fault injection system, and fault injection tests are discussed.It is indicated that the methodology can test the key resources used in the user designed circuit with high resource coverage fraction and very low time consuming, and the methodology has good practicability.

In view of multi-channel, wide temperature range of temperature data acquisition requirements of space camera, analysis is made to the nonlinear characteristic of the NTC thermistor in temperature acquisition circuit, and the method for compensation of nonlinear thermistor sensing property is presented.Taking MF501 NTC-thermistor in design of aerospace thermal control as an example, a new kind of rational regression model of voltage-temperature data is proposed based on Kalman filtering, and simulation analysis is made to which performance parameters by Matlab programming.The experimental results show that:1) In temperature range of -10~+50 ℃, the temperature measurement accuracy by using the rational regression model is 0.134 0 ℃, the memory usage is 56 bytes, and its comprehensive performance is better than that of the traditional data acquisition and processing method;and 2) The method has strong commonality and is convenient for application, which can improve the acquisition accuracy and measurement range of the temperature data effectively with low resource consumption.

A damage identification method is proposed, which is suitable for light-weight,high-strength beam structure.Aiming at the problem of positioning interference in the frequency sensitivity method of additional mass for use in the light-weight, high-strength structural damage detection, we combined the modal vibration of the structure with the strain energy, and proposed a damage identification method, which could eliminate the damage location disturbance by using the structural strain energy.Numerical simulation and experimental verification were carried out respectively.The results show that the method has the advantages of both the frequency sensitivity method and the strain energy method in damage identification, and the damage location and damage degree can be identified more precisely.

The influence of source number estimation on application of Independent Component Analysis (ICA) methods is researched.For positive model of ICA, when the estimated source number is less than the actual source number, the positive model is changed to over-determined model, and natural gradient algorithm is implied for simulation experiment.To the case that the estimated source number is more than the actual source number, the positive model is changed to under-determined model, for which the hybrid matrix estimation algorithm based on single source detection is used for estimating the hybrid matrix, and subspace representation algorithm is employed for studying the influence on source separation.Experimental results show that, as to the problem of positive model changing to over-determined model, in the case of signal to noise ratio satisfying the requirements, the number of successfully separated source signals is usually equal to that of estimated source number.As to positive model changing to under-determined model, usually all of the source signals can be successfully separated, but there is one or more of the separated signals is the copy of the original source signal.In this case, the separated signals can be merged based on correlation analysis, and all of the original source signals can be separated successfully.The results supply a referenced for applying ICA methods.

At present, in military avionics manufacture facility, it's common that purchase orders come in small patch but diversified models, the batch production share manufacture resources with trial-manufacture and repair.However, the logistics lag in huge way.To deal with these problems, the company in this case applied lean theory in the final assembly workshop by redesigning the process layout not only in the whole building but also within every assembly line.The methodology summarized here is replicable, therefore can be taken as a reference for similar projects.

A fault-tolerant control method was proposed based on adaptive control for the quadrotor UAV in the presence of actuator fault and external disturbances.The actuator fault was added to the system model in the form of multiplicative factor, and the attitude was described with unit quaternions.Considering the under-actuated characteristic of position subsystem, the virtual control force was introduced, and the actual thrust and the desired attitude were calculated out.To the position and attitude subsystem with both actuator fault and external disturbances, an adaptive fault tolerant controller was designed to track the desired attitude.Finally, the simulation shows the effectiveness and stability of the proposed controller.