Influences of drum cutting load disturbances on dynamic response characteristics of high mechanism during height adjustments of shearers were studied. According to the model of asymmetrical valve controlled cylinder, a transfer function model of height adjustment mechanisms was established. According to the working conditions of shearer, a load model of drum cutting load equivalent to height adjustment mechanism was analyzed. Conventional PID and fuzzy PID controller were used to control the mechanisms of different drum cutting loads and traction speeds. The results show that, fuzzy PID controller has better tracking performance, and the control effectiveness is less affected by cutting impedances and traction speeds.

From the viewpoints of industrialization and technology development, the paper summarized the revolutionary history of machine tools and the development processes of NCMT, with the analysis of the kernel technologies of NCMT. The development process of NCMT in China was divided into 3 periods, initial development, consistent technical research and industrialization, and high-speed development, transformation and upgrading as well. The milestone events and achievements were introduced. The industrial scale of NCMT all over the world and the advanced NCMT technologies in several developed countries were presented, followed by the comparative analysis of respects relative with NCMT at home and abroad. Finally, the future development trends were summarized prospectively.

There was a way to detect the positions of linear motors by detecting the magnetic field informations of linear motor movers. The principle of double Hall sensor position detection was introduced, and the cause of the installation errors was analyzed herein. In order to achieve the Hall sensor position detection error compensation, a method of using a 3D Hall sensor was proposed instead of a linear Hall sensor. In order to reduce external interferences, a fuzzy-neural off-line modeling method was proposed. Finally, the correctness of the compensation method was verified through the simulations and experiments.

The levels of end-toothed disc indexing accuracy might have direct effects on the indexing accuracy of the tool rests. Tooth thickness wear errors and a variety of machining errors were analyzed. An indexing accuracy error model of the end-toothed discs with multiple error factors was established. On the basis of the indexing accuracy model of the end-toothed discs, a dynamic reliability sensitivity mathematical model was established by combining reliability sensitivity analysis method. The evolution rules of reliability sensitivity for each random parameters of the end-toothed discs were given. The influences of the random parameters on the reliability of the end-toothed disc indexing accuracy were analyzed. The results show that the reliability of the indexing accuracy of the end-toothed discs decreases gradually with the the loading and unloading processes. The variation tendency of the sensitivity of each design parameters at the same time is different. The sensitivity parameters should be controlled in order to improve the reliability of the indexing accuracy of the end-toothed discs.

Based on the deficiencies of Lankarani-Nikravesh contact force model, a modified contact force model was proposed, and it was verified that the modified contact force model is effective. A parallel mechanism RU-RPR with revolute joint clearances was taken as research object, and the modified contact force model and modified Coulomb friction force model were used to calculate normal contact forces and tangential contact forces, respectively. Dynamics equations were established in form of differential algebra. Baumgarte stabilization method was added and the equations were solved by the fourth order Runge-Kutta method. Then effects of different friction coefficients on dynamic behaviours of the mechanisms were analyzed.

The development of productivity and the innovation of engineering technology required mechanisms to have the multi-function of self-reorganization and reconfiguration to meet the needs of complex working conditions. Reconfigurable mechanisms had variable mobility and varying configurations to meet the requirements of multiple tasks, multiple working conditions and multiple functions. However, the research history of mechanism evolution and furcation principle, which determined the design method, was still not fully understood by scholars. Herein, from the point of view of the evolution and furcation principle of metamorphic mechanisms, based on screw theory, Lie group, Lie algebra and differential manifold, the mechanism evolution and the interrelationship between the mechanism motions and the constraint spaces were revealed. Then, the furcation principle and the controllable singularity configuration in the evolution of the mechanism were explored. Moreover, the historical relationship between metamorphic mechanisms, origami mechanisms and deployable mechanisms was discussed, and the configuration design, performance synthesis, new design concept of metamorphic mechanisms and their innovative application were reviewed.

A method for evaluating handling comfort of vehicle gear laver was proposed based on its force feedback characteristics. The method combined the objective test with the subjective evaluation, to compensate for the inadequacy of subjective evaluation only. The variation of handling force feedback with the displacement of the gear lever and the driver's individual differences were analyzed. The relative force feedback maxima and relative handling stiffness were extracted as objective indicators for the processes of gear selection, in-gear and out-of-gear. In addition, back propagation (BP) neural network was improved by particle swarm optimization (PSO), thereby an evaluation model was established, which described the mapping relationships between objective indicators and subjective scores. Finally, the model was trained and tested by 48 groups of sample data obtained from the handling experiments, which took the manual transmission used widely in normal car for examples. The results show that the evaluation with proposed  method is accurate and stable, so that it can provide guidance for optimization design of vehicle transmissions.

Under the functions together with various kinds of uncertainty factors and time factors in the service environments, strength degradation and cumulative damages on structures would be conducted, and thereby seriously affected the structural reliability. The traditional reliability model could not reflect the influences of loading sequence and strength degradation. The problems of the multi loads equivalent transformation were studied herein by using interval theory and considering nonlinear evolution phenomena of fatigue damages caused by the different loading sequences. Aiming at the strength degradation structures under the multi-level loads, the dynamic non-probabilistic reliability prediction model was established and the analytical method of time-dependent reliability was constructed based on an interval model. Finally, by calculating an example, the time-dependent reliability for structures under multi-level loads was analyzed, and the validities of the method and model were verified.

Based on infrared thermograph technique, the laser welding processes of 304 stainless steels were recorded to study the influences of welding parameters on temperature field characteristics. The results show that, when the laser power is constant, the lengths and the widths of the melting zone are increased with the decrease of the welding speeds, but the basic characteristic of the temperature field distribution will not be changed. When welding speed is constant, with the increases of laser power, the temperature of the welding seam is increased ,especially the temperature and area of sub-high temperature are increased significantly, but the distribution of temperature is invariable. When the energy input is close to each other,the “keyhole” temperature of the high power/high speed combination is higher. Meanwhile, the temperature and area of the sub-high temperature zone in the combination of low power/low speed is higher, but it has a lower temperature at the back of the weld than the combination of high power/high speed.

The development of high-speed rail welding equipment at home and abroad was analyzed herein. The characteristics and advantages of MF inverter DC welding technology were summarized, and the mechanical structures, inverter transformers and control systems of the self-developed MF inverter DC welding machine and the product tests were introduced. The results show that the MF inverter DC welding machine developed by the project has the characteristics of dense welding currents, no obvious skin effects and high drop rates. The relevant performances are the same as that of abroad advanced levels, which have important reference significances for the welding of high-speed rail seamless rails.

As prediction accuracy was not high, and deviations were large for finite element simulations and tests, a new door stiffness analysis method was proposed, which considered test fixtures. Using theory of limit error synthesis, experimental error model was established by coupling test data collection errors and flexible deformation errors. A method was gave based on sensitivity analysis to calculate importances of error effects considering different station fixtures, and the common form of simulation error transfer model was deducted.  Finally, results show that the fixture has a great influence on deviations between tests and simulations,and verifies accuracy of the error model in an example of a car door.

To obtain the performance and design method of carrier-based arresting device,the components and the function of carrier-based arresting device were analyzed.A dynamic model of carrier-based arresting device was established and the aircraft arresting force was deduced.A set of curves about constant runout control valve area with different aircraft masses were calculated under  the constant runout conditions,also the longitudinal dynamic performances of aircraft were obtained.The results show that different mass aircrafts can be stopped in the limited distance by changing the area of constant runout control valve of carrier-cased arresting device, the overload and arresting distance are satisfied with the requirements of American military standard.

According to the structure and movement characteristics of five-axis automatic drilling and riveting systems, the off-line programming system architecture of automatic drilling and riveting was constructed, and the motion simulation module was developed. Firstly, considering that the characteristics of all points were relatively fixed in bracket coordinate system, an inverse kinematics solution was proposed based on the moving coordinate system's point positions. Based on this, a real-time interference detection route planing algorithm was proposed.Then, motion simulations were realized in DMU module of CATIA by CAA secondary development method. Lastly, an example of aircraft panel part was given to demonstrate the feasibility of this system. The results indicate that it may effectively identify interference parts and avoid interference by inserting viaing points.

Based on elliptic ultrasonic power cutting driving device of the BUCK converter model optimization research, fully considering the influence of each component parasitic parameters on the model, at the same time the LC optimization design was carried on the thorough analysis, an optimized converter model was designed and the appropriate verification circuit was made. Through analyses of two important BODE diagrams of transfer function model test of the model and simulation analysis, the results prove that the model better reflects the dynamic performance of the real circuit, better guides the design of the converter.

To establish the dynamics model of complex rotor systems in turboshaft engines,the complex structure was modeled by substructure method on the basis of finite element method and piecewise linear fitting. After reducing the degrees of freedom of the systems,motion equations of the rotor systems were derived. The validity of the model was verified by theoretical analysis and experiments,and the vibration characteristics of the gas generator rotors were analyzed. The results show that the established model may significantly reduce the complexity of the overall systems and greatly shorten the calculation time under the premise of ensuring the solution accuracy.

Considering the physical significance of crack formations and propagations under multiaxial loading conditions, a multiaxial equivalent linear fatigue life prediction model was proposed based on Miner rule and critical plane method. The ratio of cyclic yield stress to static yield stress was used to reflect the cyclic strengthening capacity of the materials, and the additional hardening factor was defined. The influences of phase differences and loading conditions on the non-proportional additional hardening effects were considered, and the mean strains under asymmetrical loadings were corrected. Multiaxial fatigue lifes of smooth and notched samples for five kind of materials were estimated by equivalent strain prediction model, maximum shear strain amplitude model and the proposed model, and the estimated results were compared with the testing results. The results show that the three models may achieve good effectiveness under proportional loadings, but the proposed model performs better than that of the other two models under non-proportional loadings. Meanwhile, it also verified that the proposed additional hardening factor may reflect the influences of phase differences and material properties on additional hardening effects.

Aiming at the problem that was difficult to obtain a high accurate breakout prediction model of continuous casting in the case of small sample data, a breakout prediction algorithm was proposed based on active learning GASVM classifier. Firstly, the algorithm preprocessed temperature data of continuous casting mold and labels valid data. Secondly, SVM model was obtained after SVM empirical parameters were optimized using labeled small sample data and GA. Finally, the optimized SVM model was tested using the historical data of a steel plant. The results show that in the case of small sample data for training model, the breakout prediction algorithm based on active learning GASVM classifier can obtain higher breakout prediction accuracy and 100% reported ratio. The presented breakout steel prediction algorithm was validated.

To realize the requirements of a certain optical instrument for flexural hinges,a new type of cross-spring flexural hinge was proposed,and the design method was studied by using the Castigliano's second theorem.First,the Castigliano's second theorem was used to derive the flexibility calculation formula of the cross-spring flexural hinges to determine the axial rigidity and rotational rigidity of the hinges.Then,a case design was carried out and the finite element analysis was used for simulation.Finally,an optical test platform was set up and the rotation stiffness of the example was measured.The results show that the analytical solution,the simulation solution and the experimental data are in good agreement,and the relative error is as 8.7%.Using the Castigliano's second theorem as a design tool,the designer may determine the dimensional parameters according to the rigidity and structural requirements of the cross-spring flexural hinges,and the design of the cross-spring flexural hinges provides a new approach method for other hinges.

Aiming at the problems of wrinkling, cracking, uneven thickness, and large die contact force in the first deep drawing processes of a power battery case, based on entropy weight a comprehensive evaluation method was proposed to optimize the processing parameters of the first deep drawing processes, so as to improve the quality of the first deep drawing processes. Orthogonal experimental simulation was carried out with the influencing factors such as blank-holder force, die radius, punch radius, die clearance, and friction factor, and the evaluation index such as maximum reduction ratio, maximum thickness ratio, maximum limit punch contact force, and maximum differential thickness. Based on the applications of entropy value method and the comprehensive evaluation method, the optimum technological parameters combination was obtained, and the forming quality of the first passes of the power battery cases is improved.

Aiming at the problems in assessing the risk of failure modes contained actual greater risk in current FMEA effectively, a new FMEA method was proposed based on rough set theory and quality loss function. First, on the basis of potential failure mode analysis, the matrix model of failure modes risk assessment was constructed by expert system. Second, in order to reduce the subjectivity and vagueness of the data itself, the matrix model of risk assessment was converted into rough domain matrix model based on the rough set theory. Third, combined with the solved rough domain matrix, potential loss of each failure mode was solved by applying multivariate quality loss function model. Risk level of failure mode could be sorted according to the potential loss. Finally, an engine assembly process was used as an example to analyze the failure modes and influences. Reasonableness and effectiveness of the proposed method were validated by comparing the results with the current FMEA method.

An overview of the development of reliability engineering was provide, and the recent developments of reliability test and problems with reliability test of mechanical products were summarized. The developments of methods on how to conduct statistical analysis and design test plan of accelerated life test and accelerated degradation test were reviewed, and also the developments of high accelerated life test and high accelerated stress screening test were reviewed. The development prospect of reliability test technology of mechanical products was presented and the guidelines were provided to promote the engineering applications of reliability test theory and methods.

考虑上配重对洗衣机悬挂系统振动的影响,基于拉格朗日方程,对该系统进行动力学建模。进行了多工况下的洗衣机筒体振动测量实验,以实验得到的筒体振幅随转速变化的关系曲线为目标,对数学模型进行了参数修正。通过数值仿真与实验结果的对比分析,验证了动力学模型的正确性。
Considering the effects of top balancing weights of washing machine suspension system, a dynamics model of this system was set up with Lagrange equations. Experiments were designed to measure vibrations of washing machine drum under multiple working conditions. An modification for parameters of dynamics model was performed based on the relation curves between the vibration amplitudes and speeds of the drum. The results calculated by dynamics model were compared with those obtained by experiments. The validity of dynamics model was verified.
washing machine; horizontal-axis; suspension system; top balancing weight; dynamics model; vibration measurement

In order to make the intelligent vehicle can run on the large angle curve road smoothly,a new curvature-continuous obstacle avoidance trajectory planning method was introduced.The trajectory planning method was based on  β-spline and utilized  the convex property of β-spline.The effection of several nonholonomic constrained conditions to trajectory planning  was considered in this method.The nonholonomic constraints included obstacle avoidance constraint,curvature constraint,rotational speed constraint,speed constraint,state constraint and so on.An intelligent vehicle's right-angle curve road running trajectory was generated.The trajectory was obstacle avoidance and the curvature was continuous.The intelligent vehicle's right-angle curve road running experiments were done to test the trajectory planning method.The results show that this method is more effective and feasible than the traditional method.Using this method curvature-continuous trajectory can be generated and making the
vehicle run along the trajectory smoothly.

This paper proposed one solution using elastic deformation of thin-plate actuated by hydraulic pressure in order to compensate the gap between ram and saddle of gantry type milling machine Z axis which might lead to bad stability and consistency during machining. Compensators with different materials or dimensions were designed according to one certain CNC machine. Performance of these compensators was researched both in theory and experiments, the outcome of which shows a linear correlation between deformation and pressure as well as negligible hysteresis for each specific compensator while varies among ones with different materials or dimensions.

A decentralized fault-tolerant control method was designed based on non-singular terminal sliding mode for a space robots subjected to partial losses of actuator effectiveness. Firstly, the dynamics equation of the system was established by using linear momentum conservation law and Lagrangian method. Subsequently, the system was decentralized into several subsystems based on the local informations of the carrier and joints, and the dynamics equations of subsystems were obtained. Then the effective factors were separated from the subsystems' dynamic equations, and the adaptive decentralized neural network was used to estimate the separated variables in real time. According to the estimation results, the corresponding control law was designed online to counteract the influences of the actuator faults on the stability of system and ensure good trajectory tracking performance. The asymptotic stability of the whole closed-loop system may be guaranteed under the proposed control scheme via Lyapunov function method. The simulation results demonstrate the effectiveness of the proposed control method.

A trajectory planning algorithm was proposed based on pose error feedback for a three-arm space robot.The pose error problems might be solved when the robot went through a singular region. The kinematics model of the three-arm space robot was established by general kinematics equation and law of momentum conservation.The pose error kinematics equations were established by position and at-titude expectations.Taking joint angular velocity as the control variable, a controller was designed based on error feedback of end-effectors and bases.The tracking errors of closed-loop system possessed speci-fied exponential convergence rates.The method may reduce pose errors in the neighborhood of singularity points and eliminate the errors completely after leaving singular regions. Simulation results demonstrate the effectiveness of the proposed trajectory planning method.

The metal powder bed fusion was an advanced and promising manufacturing technique which was rapidly developed in recent years. The extensive attentions in the fields of aerospace,biomedicine and automotive were obtained,but shortcomings such as poor process repeatability and insufficient stability were still restricting its further industrial applications. On-line monitoring provided an effective solution for these obstacles by monitoring process statuses and defects in time. The research statuses of on-line monitoring for metal powder bed fusion were reviewed from the aspects of powder recoating monitoring,powder bed inspection,building process monitoring and deposited layer inspection. Then,limitations of present online monitoring techniques were pointed out,and some topics to be further studied were predicted.

Firstly,this paper gave an overview of the application status of the angular contact ball bearings in motorized spindles and analysed the interrelationship between the preloading forces and the performance such as precision,spindle speed,service life of motorized spindles. Second,were the princi-ples and characteristics of the bearing positioning preloading technology and the constant pressure pre-loading technology reviewed. Then the adjusting preload technology and the automatic preload technolo-gy were commented with different preloading methods respectively,such as hydraulic preload equip-ment,shape memory alloy(SMA)spring,double metal sleeve,piezoelectric equipment,electromagnet-ic force equipment and centrifugal force equipment etc.. Finally,the future development trends of motor-ized spindle preloading technology was prospected.

For discrete ground contact,there were high adaptability and extensive application prospect for mammalian legged robot when encountering special topographical and unpredictable environment such
as obstacles,ditch and so on.This paper surveyed the development process of mammalian legged robots,and mainly introduced overseas and domestic research status in terms of two legged and four legged robots.The related theory and method on legged structure,joint drive,navigation,stability criterion and control algorithm  were analyzed and discussed,Meanwhile,the difficulties to be solved under development were analyzed and the future of legged robots was forecasted.

In order to realize the precise measurement of  the nN-level micro-force in the industrial technological field,a kind of spherical film,with about 700nm thick and around 50mm in diameter,was used as sensitive component.Distort will be formed when nN-level particles were put on the spherical film.According to the deformation of spherical film,we can get the gravity of the particle.By using the ANSYS software,a mechanics model of the spherical film was built.Adopting the orthogonal experimental method to design some experimental situations,then mechanics analysis of the spherical film was processed in these situations.
And then the experimental results were manipulated by using the least square fitting algorithm that was based on the singular value decomposition.Furthermore,fitting formula was obtained to describe the relationship of deformation,the parameters of spherical film,and the gravity of particles.Comparing the experimental data from situations and the calculating data from fitting formula,we can conclude that when controlling the force in the range of 2~10μN the detecting error will be less than 1%.

Three technical implementation routes of the dual-power 48 grinding stone rail grinding trains were analyzed, and the technical route of the trains adopted by "diesel generator set + overhead contact system + AC drive" was determined, then the key technologies of the trains were listed. The main structures and performances of the trains were introduced, and the traction characteristics were calculated and analyzed under the self-running conditions and grinding working conditions respectively. Finally, the railway adaptability, product advantages, social benefits and test conditions of the trains were described. The calculation analysis and test results show that the traction characteristics of dual-power 48 grinding stone rail grinding trains meet the working requirements on long steep grades, and the trains adapt to the existing passenger line of various speed grades, the strength of car bodys and bogies, the dynamics performance and the working performances all meet the requirements of test standards. The trains also have the advantages of energy saving and environmental protection, and are suitable for tunnel grinding.

Aiming at the problems that the available fault data of rolling bearing fault diagnosis were scarce in industrial productions, a multi-source transfer learning bearing fault diagnosis method was proposed based on shapelets time series. Firstly, source domain sets were constructed by the laboratory data which included abundant typical fault information and sufficient label information, and the source domain feature extractor and classifier were trained using the training data of each source domain. Then, the shapelets learning algorithm based on dynamic time warping (DTW) was used to extract the shapelets of the source domain and the target domain as the discriminant structure, the source domain data was optimized through the measurement discriminant structure, and the source domain network classifier was fine tuned to obtain the diagnostic model. Finally, according to the difference between the shapelets of each source domain and the target domain, the results of each classifier were aggregated by using the adaptive domain weight to obtain the diagnosis results. Experimental results show that the proposed method has good fault diagnosis performance in the case of few shot and high noise.

The reasonable scheduling scheme of employee-task in the automobile prototype workshop was the key to ensure the smooth progress of production. In view of the characteristics and existing problems of automobile prototype workshops， an evaluation system of employee work ability was designed based on decision-making trial and evaluation laboratory-analytic network process(DEMATEL-ANP) method. On account of the quantitative assessment of employee work ability， a workshop scheduling optimization model was established considering employee work ability， aiming at the makespan， the balance degree of skill utilization and labor cost. A non-dominated sorting genetic algorithm-Ⅱ(NSGA-Ⅱ) based on technique for order preference by similarity to ideal solution(TOPSIS) was adopted to solve the model and get the relative optimal solution. Finally， the feasibility and effectiveness of the proposed model and method were verified through practical production cases.

In order to improve the positioning accuracy of the cable-driven continuum robots, an error calibration and compensation method was proposed for the cable-driven continuum robot modules(CDCRMs). The product-of-exponential(POE) formula was employed to develop the kinematics model for the CDCRMs. The kinematics error transfer model was derived through the kinematics model, which was used to identify the kinematics errors by using the least-squared algorithm. Identified errors were used to compensate the kinematics model, thus the accuracy of the CDCRMs was improved. A prototype of a CDCRM with a flexible backbone was made to verify the calibration algorithm by simulation and experiments. The results show that the position accuracy of the CDCRMs is improved by 32.23% and the orientation accuracy is improved by 81.64%, which prove the effectiveness of the calibration algorithm.

Aiming at the problems of extracting and separating fault modulation characteristics of rolling bearings from vibration signals of a gearbox under variable rotate speeds, a new fault diagnosis method for rolling bearings was proposed based on time-varying zero-phase filter. The gear mesh frequency was estimated from the vibration signals of faulted rolling bearings in gearbox by using chirplet path pursuit, and thus the shaft rotating speed of gearbox might be obtained by dividing the gear meshing frequency by the number of gear teeth, meanwhile, the analysis of Hilbert envelope demodulation was carried out to obtain the envelope signal of the vibration signals of faulted rolling bearings. And according to the shaft rotating speeds, each time-varying zero-phase filter was designed. Then each modulation signal was acquired by the analysis of time-varying zero-phase filter to the envelope signal. Lastly, the fault diagnosis of rolling bearings was carried out according to the order spectrum, which was obtained by using order analysis to each modulation signal on the basis of the shaft rotating speeds. Simulations and application examples indicate that the proposed method may effectively extract and separate each fault modulation characteristics of rolling bearings from a rotating speed changing gearbox.

When the automobile exhaust thermoelectric generator system was connected to the exhaust pipe of the engine, the airflow would be changed. In order to minimize the impacts on the original performance of the engines and improve the electrical power generation performance of the automobile exhaust thermoelectric generator systems, a hexagonal heat exchanger used for an automobile exhaust thermoelectric generator system was optimized for the on-board compatibility with the goal of the highest average surface temperature and the smallest pressure loss. A finite element simulation model of the heat exchanger was built based on CFD software to study the influences of the inlet gas speed of the heat exchanger on the surface temperature and pressure loss, as well as the influences of the inlet gas temperature on the surface temperature. The fins length, fins angle, fins width and fins spacing were taken as the design variables. The Gaussian process regression proxy model was established based on the test data, and the grey wolf algorithm was used to obtain the optimal solution in the multi-objective optimization function space. The results show that compared with the classical NSGA-Ⅱ algorithm, the Pareto solution set obtained by multi-objective grey wolf algorithm is more concentrated and the evaluation index is higher. The optimized heat exchanger with multi-objective grey wolf algorithm has a lower surface average temperature than that of several structures before optimization, but the pressure loss is significantly reduced. Compared with the cavity structure heat exchanger, the surface average temperature and pressure loss of optimized heat exchanger increase, and the pressure loss increasing amplitudes are within the acceptable ranges. The optimized heat exchanger may effectively take into account the power generation of the automobile exhaust thermoelectric generator systems and the on-board compatibility.

导语：近年来，数字技术快速发展并得到广泛应用，在给人们带来便利的同时，也产生了流程僵化等问题,尤其在应对突发事件时，更凸显了流程僵化导致的系统的脆弱性。由此，加快从刚性自动化向柔性自主性的转变，实现以人为本的柔性设计，并构建面向未来智造的人-系统集成系统成为当下的研究重点。《Design for Flexibility：a Human Systems Integration Approach》一书阐述了柔性设计的概念、方法和过程，提出社会-技术系统柔性分析框架，介绍了人-系统集成(Human-system integration, HSI)这一新兴学科，为处理突发事件并提出问题解决方案提供了工具和方法，相应研究成果有望为提高设计和操作柔性、促进人机交互效率、构建可持续发展的社会-技术系统提供理论支撑。HSI作为一门新学科，旨在将以人为本的设计理念融入复杂系统工程的全生命周期过程，为以人为本的柔性设计提供了方法论，同时也为人本设计、人本智造等研究提供了参考。0 引言近年来数字化、网络化、智能化等技术迅猛发展，人类生产和生活自动化水平不断提高，数学技术给世人带来诸多便利的同时，也产生了流程僵化等问题。尤其在面对不可预见的突发事件时，传统管理方式面临严峻挑战，也更加凸显了流程僵化导致的系统的脆弱性。“如何管理突发事件，实现从刚性自动化向柔性自主性的过渡，构建一个更加注重柔性和可持续发展的世界”引起了人们的关注。如何合理利用新一轮科技革命技术成果，构建人-社会集成系统和可持续发展模式引起了学者Gug André BOY的思考，进而他提出了以人为本柔性设计的新理论和技术体系来解决当前的社会技术问题，并撰写了《Design for Flexibility：a Human Systems Integration Approach》^[1]一书。该书围绕以人为本的柔性设计主题，介绍了社会-技术系统柔性分析的概念、框架、模型，对进行柔性分析的人-系统集成(Human-system integration , HSI)方法展开了详细阐述，并探讨了复杂系统的有形性问题，给出了以人为本的虚拟设计(virtual human-centered design, VHCD)实体化的具体方法和有形度量指标，为设计和制造人-自然-社会和谐的人工制品提供理论指导。HSI作为新兴学科，是心理学、社会科学、生物学、数学、计算机科学等多学科交叉的一门学科，该学科所提出的基本框架可用于分析复杂问题、获得解决方案和提高社会-技术系统的柔性^[2]。人-系统集成将以人为本的设计与复杂系统工程有机结合起来，依托虚拟样机，利用人在回路仿真(Human-in-the-loop simulations, HITLS)和数字孪生技术，考虑社会-技术系统中人的因素和组织因素，实现设计过程的人机交互和产品全生命周期的仿真，促进以技术为中心的传统系统工程向以人为中心的数字工程的转变，旨在提高整个系统全生命周期的柔性。 1 复杂社会-技术系统该书第一章以新冠肺炎疫情的爆发为背景，揭示了打破传统僵化的程序、灵活处理突发事件的重要性，强调了研究技术、组织和人员方面的柔性的迫切需求，进而引出人-系统集成这一多学科交叉的新兴学科，以实现从刚性自动化到柔性自主化的转变，提高社会-技术系统的柔性。第二章介绍了社会-技术系统的柔性分析框架。考虑技术成熟度、实践成熟度和社会成熟度三种与自主性密切相关的成熟度以及管理复杂系统的三个主要过程(图1)，阐述如何实现从刚性自动化到柔性自主化的转变，这个转变需要定义“系统”的一致表示，即系统是由结构和功能组成的，提出图2所示的“情境-资源正交”框架。书中对资源和情境的概念分别进行了定义，其中资源被形式化为客体或主体，情境由结构、功能和动态三种因素定义，与TOP模型(technology, organization and people)(图3)结合构成情境分析全局框架，如图4所示，最后区分情境与情景的概念，提出态势感知(situation awareness, SA)的情境模型框架，如图5所示。图1 程序执行、自动监督和问题解决过程图2 “情境-资源正交”框架图3 TOP模型图4 HSI情境框架图5 态势感知的情境模型第三章介绍了实现柔性设计的基础方法和模型。首先对复杂社会-技术系统进行定义并对复杂系统的特性进行阐述，引出SFAC模型(structure/function vs. abstract/concrete)(如图6所示，提供了人工制品的结构和功能之间的等效表达，即抽象和具体)、NAIR模型(natural/artificial vs. cognitive/physical)(如图7所示，阐述了社会-技术系统的认知功能和物理功能的区别)和AUTOS金字塔模型(artifact, user, task, organization and situation)(如图8所示，是TOP模型的扩展，一个简化的HCD工程框架)，图6 SFAC模型图7 NAIR模型图8 AUTOS金字塔构建以人为本的柔性设计的概念框架^[3]，并分析系统的复杂性。其中，复杂系统的特性主要包括：①组件及组件间的互连；②多人参与全生命周期过程；③未包含在组件中的全局属性或行为；④复杂适应机制和行为；⑤不可预测性的表征。 2 人-系统集成(HSI)第四章详细阐述了人-系统集成的具体内涵。介绍了人-系统集成的认识论及演变，说明可分离性问题是复杂系统的一个重要特性；阐述了HSI的仿真方法--人在回路仿真(HITLS)，并分析在系统全生命周期考虑资源投入、设计柔性和系统知识三个参数的影响；分析了从传统工程到数字工程的转变，回答了如何提供更多的自主性和柔性问题(图9)；引出在这个过程中获取程序性和陈述性知识的工具(人工智能)和方法(PRODEC方法)。图9 从传统工程到数字工程的转变-- 从刚性自动化到柔性自主性的转变第五章介绍了基于活动的设计。通过基于场景的设计(表现形式如图10所示)处理任务分析，通过人在回路仿真完成活动观察和分析，进而实现基于活动的设计。分析了从HighTech(最先进的可用技术)到FlexTech(支持工程设计和操作柔性的技术)的演变过程，包括从人因和人机工程学(HFE)、人机交互(HCI)和图10 人机系统形成性评价的多规则方法人-系统集成(HSI)三个时期，如图11所示，这种演变创新被视为一种冒险活动。作者从互联网快速流行的原因、航空公司意外事件的管理，证明了创新需要冒险。图11 以工程为导向到以人为本设计的演变第六章分析了基于模型的人-系统集成及其柔性。首先介绍了影响HSI柔性的重要因素：态势感知、具体化和熟悉度；接着给出了预测模型和知识模型两种模型，并以新冠肺炎疫情为背景，说明了两种模型的作用，证明不同模型只在特定情境中有效，提出基于经验的建模方法，在这个过程中，识别突发情况至关重要；最后介绍了监督、调解、合作三种系统交互模型(表1)，以支持更多的自主性、协调性和柔性。 表1 系统交互模型