Aiming at the problem that the traditional eight-beam ratio amplitude direction-finding system has huge calibration workload and high real-time requirement， this paper proposes a method to calibrate the system using only five angle points per angle interval， and uses field programmable gate array （FPGA） devices to complete the calibration according to the real-time requirement of the direction-finding system. The whole direction-finding system is divided into internal and external channels， and the system correction source is introduced into the internal channel to realize the dynamic correction of the direction-finding channel， reduce the system error and improve the maintainability of the system. Field test shows that the calibration method can reduce the testing workload and the final system direction finding accuracy meets the requirements of engineering application. The root mean square error of all band direction finding is 2.01°， and the real-time requirement can be achieved by using FPGA device.

Frequency shift interference can produce high fidelity false targets for the range-Doppler coupling characteristic of LFM signal. According to the peak characteristics of pulse echo by compression and the principle of dechirp， the interference suppression method based on signal reconstruction and cancellation is proposed. Firstly， the frequency of the signal by dechirp is estimated， and the dechirp interference signal is reconstructed. Secondly， the normalized average detection threshold is defined for the dechirp signal， and the minimum threshold is used as the objective function of interference suppression. Finally， the interference suppression factors corresponding to the minimum normalized average monitoring threshold are calculated by global search， and interference suppression is realized by cancellation. The simulation shows that the energy of residual interference component is more than 20 dB lower than that of target echo in the output signals by pulse compression， which shows the proposed method can suppress multi-frequency shift interference effectively.

With the increase of the level of complexity， integration and intelligence of missile weapon system， health management technology， as a key enabling technology to improve equipment reliability， testability， maintainability and supportability， has gotten more and more attention and application. Based on the health management needs of the new generation missile weapon system， the design principle， structural framework， workflow and life cycle data management requirements of missile equipment health management system are introduced. The feasibility and effectiveness of the design method in the development of health management system are verified by taking certain equipment as an example.

Aiming at the process of transition between high orbit early warning satellites and long range early warning radars in the course of missile defense, two issues were investigated: ballistic error propagation and missile warning information representation. The calculation process of ballistic error propagation using CADET was described. For the results of the error propagation, four kinds of early warning information models, which included the probability ellipsoid model, the probability spherical model, the uniform ellipsoid model, and the uniform sphere model were proposed. The missile warning information of different granularities was calculated using the four models. The simulation results showed that the early warning information models established described the missile state more fully.

Aiming at the conflict and uncertainty of the test data required in the quality assessment, and the incompleteness and inaccuracy of the condition assessment, an improved method of evidential reasoning (ER) is proposed. After the quality grading and test data pretreatment, the conflict between the information and the overall uncertainty are assigned to each level of probability to reduce conflict and overall uncertainty between the evidence assessed. The method is applied to the missile equipment quality condition assessment, and the results and algorithm comparison show that the improved evidence reasoning method can effectively reduce the conflict of evidence, and the overall uncertainty of condition assessment is greatly reduced. The principle of the method is simple, the theoretical basis is sufficient, and the program is convenient to be realized.

The economic and efficient level of repair is needed for the invalid parts in reality, which is often affected by many factors. Combined with the existing results, the decision-making flow chart is utilized for multi-level maintenance process of a certain equipment from the perspective of qualitative and quantitative. A multi-objective optimization model including economic and non-economic indicators is built. The optimal solution is determined by the combination of TOPSIS (technique for order preference by similarity to an ideal solution) and genetic algorithm which is illustrated effectively, and provides a reference for maintenance support of equipment.