Abstract:Based on Udwadia-Kalaba theory，a method for calculating the dynamic load of suspension part was studied by taking the six-component force of wheel center as the input. Taking the five-rod rear suspension as the modeling object，the dynamic model of the suspension system was established，and the analytical expression of the load at each part was obtained with the decomposition of the total binding force. Compared with the traditional Lagrange modeling method，the method does not need the Lagrange multiplier and reduces the complexity of modeling. In order to verify the correctness of the method，the numerical results were compared with the simulation results of the traditional modeling software Adams/Car. The analysis result shows that the proposed method is accurate and efficient，and the calculated dynamic loads are in good agreement with the Adams/Car simulation results. Finally，the method was applied to the actual vehicle test，and the load prediction for the rear suspension of a SUV was carried out.

Abstract:To efficiently solve the multi-crane scheduling problem during the production process，a novel projection model of trajectories the proposed with the consideration of initial states of tasks and space-time constraints. Based on the differential evolution algorithm，stock allocation rules and a crane allocation algorithm were combined to guide the optimization process in each iteration. Taking the minimization of delay cost and waiting cost as the evaluation indices，the simulation experiment was designed and compared with the classical multi-objective optimization algorithms. The results show that the algorithm is effective and feasible. Further numerical experiments indicate that reasonable scheduling rules can effectively improve the crane scheduling performance.

Abstract:In order to optimize the ride comfort of the sliding door under the condition of meeting the handiness requirement，a respond surface method-based multi-object optimization method of the sliding door′s dynamic characteristics was proposed. The rigid-flexible coupled multi-body dynamic model of a MPV′s sliding door was established to investigate its ride comfort and handiness，and the results were verified by experiment based on a test rig. The motion mechanisms of the sliding door system were parametrically modeled and combined with the Latin hypercube sampling method. The sensitivity analysis was performed to study the effects of structural parameters on the sliding door's dynamics characteristics. Second-order response surface models of the dynamic indexes were established，in which correlation coefficients R2 are above 0.9. Finally，a multi-objective dynamic optimization on the sliding door was conducted by using NSGA-II algorithm. After optimization，the peak values of the middle guide roller load and acceleration of the door center mass were reduced by 39% and 24.4% respectively，and the time required for full opening was below 1.8 second. The ride comfort of the sliding door was improved significantly，while the requirement of handiness was met.

Abstract:This paper studied the problem of low precision for variable pump displacement control method of closed hydraulic circuit. The linear relationship between the duty cycle signal and displacement of pump was derived based on the pulse width modulation（PWM） control of the servo valve for the pump control system. The simulation model of the pump control system was built on the AMESim software platform，and the model was verifiedby comparing with the test data. Based on the pump control system model，the PID feedback control，feedforward + feedback control， and three-step control of pump displacement were proposed，and an analysis was performed. Simulation result shows that the three-step control is superior to PID feedback control in dynamic response and rapid stability. When compared with the feedforward + feedback control，the dynamic response error of the pump displacement controlled by the three-step method is reduced by 35.5%.

Abstract:To improve the prediction precision of failure probability of machine parts，failure probability prediction method of generalized regression neural network based on GRA and AHP was proposed. The main influence factors on failure probability of mechanical parts were analyzed by grey relational analysis method based on the analysis of influence factors on failure probability of mechanical parts. The hierarchy model of evaluation index for failure probability of each mechanical part was constructed and the weight of each index was evaluated by analytic hierarchy process. Then，the weight and initial value of each index were combined to obtain the weighted evaluation value of each index. Finally，the generalized regression neural network was used to establish a predictive model by using weighted evaluation value of each index to predict the failure probability of mechanical parts. This optimization method was applied to predict the failure probability of upper gear disk in numerical control rotary table. The prediction results of traditional generalized regression neural network ，BP neural network and regression analysis method were compared. The result shows that the prediction error of the proposed model is less than 0.8%，and the residual error is in the range of -0.2% and 0.2%，which is better than the comparison models. Meanwhile，the model established by using the proposed method in this paper has higher accuracy and stronger stability，which is suitable for the prediction of failure probability of parts.

Abstract:For the fatigue life prediction of structure metalic W-rings in high temperature and pressure working environment，considering the loads of gas temperature，medium pressure and bolt preloading，the stress analysis and fatigue life prediction of metal seal ring were completed based on the thermal-mechanical coupling method. The results show that the position of the maximum stress is consistent with the minimum fatigue life，both at the outer surface of the wave arc. The seal ring has the highest fatigue life under the condition of gas temperature of 500 ℃ and pressure difference is 0.5 MPa. The method for predicting the fatigue life of metal seal ring is reliable，which provides the reference for the design of advanced metal seal ring.

Abstract:When studying the surface charge accumulation and dissipation law of insulating materials under HVDC transmission lines，the charge density can only be calculated by surface electrostatic potential inversion of insulating materials. However it is difficult to accurately calculate the surface charge density when the insulating material surface is curved. In this paper，a charge density inversion method for curved surface was proposed. The potential measurement points were selected reasonably and intensively，and then curved surface was divided into triangular meshes according to the selected measurement points. The normal vectors at each point were obtained，and the surface charge density was then inverted. As an example，the surface potential of 3D printed curved surface and suspended insulator was measured，and the surface charge density was calculated. The calculated results are in accordance with the theoretical analysis trend. The electrostatic potential of insulating material surface has a strong correlation with the charge density，and the charge density corresponding to the high potential area is also large. The charge density presents a“concentric circle” distribution according to the distance from the discharge electrode，but the charge near the suspended insulator cap is difficult to accumulate，and the charge density is small.

Abstract:In order to solve the problem of fast service restoration in distribution network with distributed generators （DGs），a multi-stage service restoration method is proposed. The service restoration method organically combines the four phases of service topological identification，islanding，connectivity restoration with main network of DGs and network reconfiguration，and load shedding operation. Once the distribution network is faulty，the service area is identified by the service topology identification method to determine the next stage of service restoration execution and to update the topology parameters. During the island configuration phase，the islanding method based on depth-first search algorithm is proposed，Which gives the priority to restore the key load and maximum loads as the goal and to improve the utilization of DGs. Heuristic rules to optimize the tie switch quickly restore the out-of-service areas and main network connectivity. Network reconfiguration phase based on the parent-child list rules avoids infeasible solution The flow calculation is used to verify whether or not the next phase is executed. The load shedding stage removes the non-critical load to ensure the stable work of the system. The example shows that this method is suitable for the case of large-scale out-of-service，multiple failures，and cascading failures in the distribution network with DGs. It can make an effective restoration strategy according to the actual situation and improve the efficiency of complex fault restoration with strong applicability.

Abstract:In view of the intermittent distributed power output and load variability，energy storage becomes an important way to support DC microgrid voltage and improve power quality. In order to make full use of the advantages of hybrid energy storage to achieve reasonable power distribution，an autonomous-frequency-split power control strategy for hybrid energy storage is proposed. In this method，the equivalent output impedance of converters is reshaped by introducing the virtual impedance or virtual capacitor in the droop control. By this way，the advantages of supercapacitor and battery can be integrated. At the load mutation，supercapacitor rapidly absorbs the high-frequency component of power fluctuation to improve the dynamic respond speed. Meanwhile，battery is mainly used to balance the low-frequency power in order to extend the service life and ensure the reliable operation of system. The simulation verifies the correctness of the proposed method.

Abstract:This paper studied a method for measuring the mechanical stress wave in power electronic device by means of test circuit and digital filtering technology. Time domain and frequency domain characteristic parameters of mechanical stress wave，such as amplitude，peak-to-peak，peak frequency，and frequency range，were obtained through signal processing and spectrum analysis. The research results show that the mechanical stress wave can be measured by setting the sampling threshold and stopping the band frequency reasonably. When a current of 40 A in IKW40T120 IGBT device is turned off，the mechanical stress wave continues to decay for 100 μs，and its amplitude and peak-to-peak value are 5.2 mV and 9.6 mV，respectively. The amplitude spectrum clearly has three frequency segments that are 20~100 kHz，150~200 kHz and 290~310 kHz. Each frequency segment has one peak frequency point，which is 54 kHz，163 kHz and 299 kHz，respectively，showing almost 1，3 and 5 octave relationship. The peaks corresponding to three peak frequency points differ greatly，which are 1.24 mV，0.69 mV and 0.36 mV，respectively.

Abstract:In order to optimize the implementation delay and hardware resource consumption of the Coordinate Rotation Digital Computer（CORDIC） in pipeline architecture，a new implementation method based on look-up table was proposed，which completely eliminatesd the iterative computation. This method requires only a lower capacity ROM table，as well as a simple shift operation on the output of the ROM table，so that sine wave or cosine wave output with high accuracy can be obtained. Theoretical simulation and practical verification were carried out in Matlab，Modelsim and XILINX ISE，and the results show that this method of CORDIC only requires two clock cycles of processing delay，and the hardware resource consumption is also reduced when compared with other methods. Additionally，the circuit output accuracy and maximum working frequency are also improved at a certain level.

Abstract:According to the actual application of a standard library cell operating in the near-threshold voltage region，and due to the problem of large error in the lookup table of traditional library files，this paper proposed a method to characterize the standard cell in near-threshold voltage region. The method redefined the boundary of the lookup table by analyzing the actual application of standard cell in near-threshold voltage，and by analyzing the relative error between the circuit synthesis result and circuit simulation result，it re-determined the scale of the lookup table，in order to improve the accuracy of standard cell library in near-threshold voltage region. This method was then used to characterize the smic55nm CMOS process library file in 0.6 V voltage and evaluate the relative error，and the results show that when compared to the library file established by traditional characterize method，the proposed method improved the library file's accuracy by 16%~63.51%，reduced the error of lookup table，and effectively improved the accuracy of library file.

Abstract:An automatic function test system for LoRa wireless tunnel lighting network devices was designed. It uses INSPEC configuration software as the development platform，and it can automatically complete function test and torture running for as many as 15 LoRa wireless network equipments at one time. It has the functions of inputting of system process parameters，real-time acquainting，calculating and processing，storaging，displaying，viewing and report form checking of test parameters，fault self-diagnosis and alarm etc. The circuit design of the test system was given，and the flow charts of function modules of the test system and the design methods and implementation principles of the key programs were described in detail. The correctness and high performance of the test system verified by the application result of nearly 10，000 sets of LoRa wireless tunnel lighting network equipment used by a company in Hunan.

Abstract:This paper proposed an improved March CS algorithm to complete the built-in self-test of SRAM memory due to the problem of March C + algorithm's fault coverage. The static and dynamic fault of memory cell were sensitized by the original algorithm increasing the read-write operation of the algorithm element， so that the fault coverage was enhanced. Finally， the March CS algorithm achievesd 91.67% and 76.93% coverage of static and dynamic faults respectively through the simulation experiments of the 1 024*32-size fault static Randon-Access memory and the measurement of FPGA to SRAM chips.

Abstract:Aiming at the leakage problem of boiler tubes in power plants，the location method based on time delay estimation was studied. The leakage signal of the furnace tube was simulated by the external high-frequency sound signal，and the multi group delay value was measured by the cross-correlation algorithm. The location equation was established according to the distance formula，and then solved by Newton iterative method，by which the leakage point coordinate was obtained. The experimental results show that the time-delay is effective only for impulsive sound source signals，and when it is the continuous periodic source，the estimation of time-delay is in the range of -T/2~T/2 （where T is the period of the source signal）. In a certain range，the higher sampling rate results in the higher delay value precision and the better effect of location；when compared with the sound speed，the sampling rate has greater impact on the location results.

Abstract:A novel low-voltage low-power ring oscillator was designed，which was implemented by an SMIC's 180 nm mixed signal processing technology. Based on the feedback theory，the regulation and conversion from supply voltage to operating voltage of ring oscillators was completed by employing an amplifier，generating superior operating voltage stability as well as lower consumption. Furthermore，a high output swing was obtained with an amplitude conversion circuit. The operating frequency of the oscillator was controlled by current，resulting in the reduced effect of oscillating signal on the supply voltage VDD and the depressed supply noise. The results demonstrate that the output frequency is 2.737 MHz，the consumption is 0.8 μW with a 1 V supply，and the phase noise at the phase point is -108.7 dB at 1 MHz offset. Meanwhile，the output frequency fluctuations are kept within 0.23% when the supply voltage varies between 0.9 V and 2.1 V，which is reasonable in passive chip design.

Abstract:Aiming at the problem of cognitive wireless network energy efficiency under underlay spectrum sharing mode，a double improved particle swarm optimization algorithm（DIPSO） for power control was proposed. The optimization of network energy efficiency was realized by recognizing the transmitted power of users under the condition of minimizing constraints. In the simulation process，under the circumstance of considering channel fading and noise interference，the fuction of network energy efficiency and relevant constraints were constructed on the premise where the normal communication of main users was not affected，the basic communication of users was guaranteed， and the transmission power of cognitive users in the cognitive wireless network was minimized. The simulation shows that the proposed algorithm can effectively improve the energy efficiency of wireless network.

Abstract:A path analysis method of the laser ignition system was investigated. First，an optical transmission model was given for the single optical path，and an analysis method was proposed by analyzing the effect of the number of optical connectors on the insertion and reflection loss. The criteria of the optical path for different number of connectors and work of laser explosives was deduced to improve the adaption and precision of the method. The experiments of ignition proves the correctness of the proposed method，which provides a reference for the design of onboard controlling computer software of launch vehicle.

Abstract:Based on the data of excellent driver's operation records，a framework of energy-saving driving optimization was proposed using machine learning technology. Hierarchical decomposition was applied to integrated machine learning method to excavate the hidden driving patterns from the driving log data of excellent drivers. The learning and forecasting of speed information and gear information were separately carried out to realize the automatic driving decision for energy-saving optimization of a train，and the actual railway lines and locomotive data were used for experimental verification. The test results show that under the constraints of ensuring the safety, punctuality and stability of the train, the energy-saving driving program of the train can save about 7% energy when compared with the average level drivers.