Volume 14    Number 7    (2002)

 

Title Author Abstract
Image Compression Technology ZHANG Hai-yan, WANG Dong-mu, 

SONG Ke-ou, GUAN Bi-gen

With the development of information network and multimedia technology, digital image, as a kind of the most important media, is used much widely. Image compression becomes more and more important for great deal of data. In this paper, some most popular compression algorithms: JPEG, Fractal compression and Wavelet compression are introduced, the advantages, disadvantages and developments of these algorithms are discussed, and some other compression algorithms are mentioned. Then an algorithm is indicated that using Polar-Log coordinate mapping as a preprocessor, which simulates the characteristic of human visual, a higher compression rate can be obtained. Finally, the conclusions and expectations of image compression algorithms are given.
Nonlinear Temperature Modeling and Prediction of MCFC Based on RBF Neural Networks SHEN Cheng, CAO Guang-yi, ZHU Xin-jian According to the drawback of the models existed which are too complicated, we set up a nonlinear temperature model of MCFC using RBF neural networks identification technology. The temperature characters of MCFC stack are briefly analyzed. A summary of RBF neural networks for the multi-input and multi-output (MIMO) nonlinear system identification is discussed, and the identification structure, algorithm and model training process are given in detail. The validity and accuracy of modeling are tested by simulations, and the comparison between RBF neural networks and BP neural networks identification is given. The simulation tests reveal the modeling method is feasible, and the training of RBF networks is much faster than BP networks. The modeling process avoids using complicated differential equation groups to describe MCFC, and the input-output performance can be quickly gotten by neuro-calculation, which makes it possible to predict and control the temperature of MCFC online.
Theory and Application of Visual Simulation Modeling for Complex Engineering System ZHONG Deng-hua, HU Cheng-shun Based on the idea of gradation and modularization presented by Zeigler, an object-oriented visual modeling method for complex engineering system is proposed. Two gradation models are constructed, whose understratum (CYCLONE) is encapsulated and hidden and can be transplanted, and superstratum (CPM) is oriented to users directly. The method is so universal that it overcomes the complexity and speciality of the modeling for engineering system. With the help of powerful VC++, corresponding simulation modeling software is drawn up, whose operation is easy and function is perfect, and it can meet different requirement of different users. The successful application of the software to practical engineering shows the effectiveness of the method.
An Object-oriented Development of Urban Traffic Simulation Laboratory System Zou Zhi-jun, Yang Dong-yuan According to the object-oriented software development concept, this paper provides an entity analysis of the urban road traffic system firstly and thereafter defines the main object classes of the urban traffic simulation system. Furthermore, the static object models which describe the static relation among object classes and the dynamic object models which represent the operating mechanism and control structure of simulation system are put forward. Finally, based on the object modeling results, the detailed object designs including the definition of objects’ attributes and functions are given.
A Study of Tracking Filter Initiation Algorithm Based on 2-Step Estimation LIU Gang, PAN Quan, ZHANG Hong-cai The design of initial gate is a crucial problem in multi-target track system. The ratio of measurement noise’s standard deviations to sample period is considered for choosing the method of filter initial in target tracking problems. Then the prior knowledge of velocity is used for constructing a ring gate of filter. Considering measurement noise for the ring gate, and the character of matrix norm, a new tracking filter initial algorithm based on 2-step estimation was developed. The algorithm is illustrated via Monte Carlo simulation where a more reliable confidence interval of gate is achieved over two other algorithms.
Fuzzy Neural Network-based Inverse Kinematics Problem of Robotic Manipulators WANG Hong-bin, SONG Zuo-shi, 

WANG Hong-rui

Fuzzy Neural Network(FNN) shows the advantages of both the Neural Network and Fuzzy Logic, which can express the fuzzy and quantilitive knowledge easily and has good learning capability. A method based on FNN is firstly proposed in this paper, which deals with the inverse kinematics problem of Robotic manipulators. The fundamental idea of the FNN and specific algorithm are given. The simulation results are presented for a 2-DOF rigid Robotic manipulators, which validate the effectiveness and feasibility of the proposed scheme.
Simulation Software of Control System Diagram Configuration Based on Elements XUE Fu-zhen, HOU Fei-li, LUO Chao This paper introduces a simulation software of control system block diagram configuration based on elements, including its functions, implementation method and improved simulation algorithm. In numerical computation, linear interpolation method of Lagrange is introduced, which decreases greatly iteration computation and preserves simulation accuracy with the original algorithm. Using the visual simulation platform, user can easily and quickly build the simulation models described by block diagrams for linear system, nonlinear system, multivariable system, multivariable system with smith predictor and sampling system etc. Hence it offers a platform of simulation study in order to analyze and design the complicated systems. This system has the advantage of friendly interface and easy development.
Design Method of Radial Basis Function Network Based on Improved Grads and Mountain Clustering ZHOU Shen, MA Zheng It is proved that the three-layer artificial forward neural network can, with a sufficient number of hidden-layer units, approximate continuous functions to any degree of accuracy. But the key issue is how we can construct these approximating networks and how many hidden-layer units we need to approximate specific functions within some specified error. To solve this problem, we propose a improved fuzzy clustering on the basis of Radial Basis Function Network’s structure to orient the position of RBFN’s central points. This fuzzy clustering method constructs density curve of the training data, sorts out the densest point which means point posses the maximum density value one by one from the gridding points defined through even plotting of training region. After that, we use nonlinear least square algorithm to find a new recursive learning algorithm which fully make use of the second order information of the RBFN to approximate the objective functions. This hybrid algorithm is proved effective in approximating a mono-peak function and the peculiarity curves of different pumping stations. We also show that the approximation rate and accuracy can meet our desired expectation.
Hybrid Flow-shop Scheduling Approach Based on Genetic Algorithm WANG Wan-liang, YAO Ming-hai, 

WU Yun-gao, WU Qi-di

Hybrid flow-shop scheduling problem (HFSP) is expansion of general flow-shop scheduling problem. It is more complex than general flow-shop scheduling problem because there are parallel machines on some operations. In this paper, a new method solving hybrid flow-shop scheduling problem based on genetic algorithm is proposed. A new coding method for the hybrid flow-shop scheduling problem is proposed. Corresponding crossover and mutation operators are given. These operators can easily keep the feasibility of solution. They can also satisfy the stochastic demand of genetic algorithm by itself. Finally, an example of production scheduling problem for metalworking workshop in a car engine plant is given. The results show the effectiveness of the algorithm.
Design and Simulation of Fuzzy Control Algorithm of Traffic Level of Service Xu Li-qun, Jiang Fu Level of Service is basically concerned with the quality of service provided by a roadway and it is very important to keep the expected level of service on road with rapidly rising traffic flow, especially under traffic congestion. A control algorithm of signal timing at intersection of link, which can control the traffic volumes into link by modifying the signal timing at intersection of link, is designed for keeping the expected traffic level of service of link by using fuzzy control theory in this paper. Simulation analysis indicates that the fuzzy control algorithm of signal timing of intersection can reflect well traffic engineers’ ideas about making decisions, and it is also proved that the signal timing must be adjusted with the dynamic traffic volumes in order to keep the expected level of service of link.
Parameter Optimization and Simulation for Fuzzy Tree Model Based on Genetic Algorithm YUE Yu-fang, MAO Jian-qin Improved from back propagation of fuzzy tree model, a kind of Takagi-Sugeno fuzzy model, this paper proposes a new parameter optimization method based on genetic algorithm (GA). Two keys are to select the way of determination for parent nodes’ α  and to select what variables as genes on chromosome. Benchmark simulations illustrate that the method have less sensitization for initial parameters and well evolution efficiency. On the same precise level, new method can simplify partition of data space and decrease computational load.
Fast Track Initiation Algorithm in Clutter Environments Based on Hough Transform and Logic WANG Guo-hong, SU Feng, MAO Shi-yi

HE You

Track initiation is the key problem of target tracking. The paper presents a new track initiation algorithm which is based on Hough transform (HT) and logic-based track initiator. In the new method, HT is used first to eliminate clutters, so as to produce a environment of low false alarm. And then tracks are initiated by using the 3/4 logic-based method. Simulation shows that the proposed method in this paper is effective in fast track initiation and is superior to HT based track initiator and logic based track initiator.
On the Bussgang Class Blind Equalization Algorithms Based on Tricepstrum Initialization Wang Feng, Zhao Jun-wei ,Li Hong-sheng, Guo Ye-cai The bandwidth of the underwater communication channel is very limited. Compared with the adaptive equalization algorithms, blind equalization needs no training sequence, so that save lots of band width, which makes it very suitable for the scenarios of high data rate underwater communication and multipoint communication. Blind equalization algorithms of Bussgang class have the advantage of computing efficiency and disadvantages of slow convergence rat sensitivity to initialization conditions. Improper initialization will lead the algorithms converge to local minimum. The polyspectra algorithms converge fast and always converge to global minimum, but the computational complexity made them unsuitable for hardware implementation. A new method based on tricepstrum initialization is discussed in this paper. According to the cepstrum –cumulant equation, the cepstrum parameters are calculated from a data set, and the tap weights estimated directly for the initialization of Bussgage class algorithm. This method accelerates the convergence rate of the Bussgang class algorithms, and ensures them converging to global minimum. Through computer simulation of shallow water channel equalization, the efficiency of the method is proved.

Genetic Optimization Algorithm of Tactical Missiles Shooting Sequence in Attacking Single Target Based on Modified Genetic Algorithm

WANG Min-le, GAO Xiao-guang Aiming at the problem of optimizing tactical missiles shooting sequence in attacking single target, the mathematical analysis of tactical missile shooting process is given, and then a modified Genetic Algorithm is presented through designing natural number encoding scheme and gene exchange-based genetic operators, so the intelligent optimization of tactical missile operational effectiveness is realized. The experiment results show that the new algorithm’s convergence speed and optimization effect have advantages over the traditional algorithms.
Change Simulink/Stateflow Model to HLA Federate TIAN Xin-hua, FENG Run-ming,

WENG Gan-fei, HUANG Ke-di

This paper presents the utility and process of changing Simulink/Stateflow model to HLA federate, and describes the principle and method which the transformation from Simulink/Stateflow model to HLA federate depends on, including object-oriented encapsulation of model, HLA-wrapper, modifying scheduler of run-time interfaces and time management of model. At last, the paper describes SMIT (Simulink Model Import Tool) which can automatically transform Simulink/Stateflow model to HLA federate, including its use-case and architecture.
Application of Agent-Oriented Method in Computer Generated Force WANG Hui-xia, WANG Xing-ren Computer Generated Force (CGF) is an intelligent autonomous entity; it has the ability of sensoring, communication and collaboration. Agent-oriented method can be used to instruct the developing of CGF system. The main issues involved in the researching and developing CGF system are the structure of agent; behaviors of sensoring, cognitive and decision-making; and multi-agent system. This paper first presents the basic concept of agent. The characters of agent to be used in computer generated force system are summarized. Some questions about designing and construct CGF system are discussed. In addition, cooperation between agents is discussed in this paper. A conceptual model of CGF based on Agent-Oriented method is presented at the end of the paper.
Simulation for a New Real-time Ethernet MAC Protocol SHEN Gang, LI Yong-nan, WEI Zhen,

XU Xiao-ming, HE Xing, ZHANG Wei-dong

OPNET is a powerful general-purpose network simulator which is widely used in protocol design, optimization and evaluation. This paper introduces a method for modeling broadcast protocols in OPNET by illustrating a simulation for a new real-time Ethernet MAC protocol. The network model is composed of nodes and links. Node models describe behaviors of nodes which are composed of interlink modules. Module functions are described by finite state machines (FSMs). The paper introduces how to construct node models and link models. The detail FSM implements of the MAC module are presented. The simulation results demonstrate that the protocol is correct and effective.
Reusing Methods in the Distributed Virtual Battlefield Simulators ZENG Liang, ZHENG Yi, LI Si-kun This paper describes reusing methods in the distributed virtual battlefield simulators. First we discuss the structural module method to design the virtual battlefield simulators, then use class module method to discuss reusing methods in the distributed virtual battlefield simulators. The samples illustrate that using the methods in this paper to design distributed virtual battlefield simulators is convenient and reusing well.
Realization and Analysis of the Air Attack Module in the Simulated Antiaircraft Training System LU Bo, HUA Zu-yao Limited by external conditions, training on the real equipment, which is costly, can only be applied in some special ways. Nevertheless the operation training can be fulfilled efficiently by means of the training simulator. To present a vivid target and simulate to electronic countermeasure which will be exist in the modern battlefield without fail is an extremely important link in simulated antiaircraft training. Based on the simulated army antiaircraft training system developed at home, this paper describes the constitution and the information exchange mechanism of the simulated air attack module in it. Besides the model method and realization technic of creating landform through 3DEM, control of the aircraft’s behavior and plan of flight course as well as the electronic countermeasure simulation are also included.
Software Design of Some Type Missile Launching Control Training Simulating System Based on Event-oriented Simulation WANG Jie, ZHANG Duo-lin, HAN Hua-ting, SHI Jian-jun On the basis of analysing simulating system function and software demand, the paper gives the software design of some type missile launching control training simulating system which adopts the simulation strategy of the event scheduling method, and event-oriented simulating method and database technology. By means of database technology, event management and save of disposal results are accomplished, the problem that a large number of signals bring about the complex logic construction of the software is solved, and dynamic database solves real-time problem of simulating system. Simulating system command, and command event list, and fault simulating event list are formed in main control program; and at same time, clock pushing is also completed, input signal and output control signal are dealt with in selecting and control program. The whole simulation process of the launching system, and the real-time supervision and replay of the operating process are realized. It is clear to the software construction, and convenient to maintenance and transplantation, running results show the event scheduling method and event-oriented simulation model are very viable and valid, so the method of designing simulation software can be used in other simulating system.
The Training Simulator of Container Crane WANG Chong-hua, LIANG Gang,

LIANG Cheng-ji 

In this paper, the background and practical significance of research and application of training simulator of port machinery are presented. And based on virtual reality and program logic control system (PLC), the new training simulator of container crane is built up. In addition, the system structure, operation principle and functions of the simulator are introduced.
A Software Mode Based Operator Training System for Hydroelectric Plant YU Tao, BAO Jie, DENG Qi-wei,

SHEN Shan-de, ZHU Shou-zhen

The paper presents a practical software-mode-based operator training system for a hydroelectric plant. The architecture of the system is briefly introduced. The modeling and simulation of a hydroelectric plant is also discussed. Afterward, the integrating methodology of the system is provided in coordination with the internal software structure of the system. The application to a hydroelectric plant in South China demonstrates the operator training system can satisfy the training needs.
Training and Simulation System of Shipborne Crane Zhang Jie, Zhang Rui, Du Yan, Li Wen-xiu Based on the synthetical study of the characteristics and work condition of a new kind of shipborne crane, this paper puts up with an overall structure of a simulation and training system for the shipborne crane. And all the component parts are simulated in a computer according to their mathematical model. It also presents the main function and characteristic of the simulation system. It is a necessary and creditable simulation model not only for the building of a pure real model but also for the further study on the characteristic of the shipborne crane. What’s more, this simulation system has a signality of directing of work on the sea and training for the operators.
Double Valves Compensation Feedback Control in Fin Stabilizer Load System LIU Zhuo-fu, LI Fu-yi, PENG Xia-fu The article analyses the characteristics and effects of redundant force in general load system, and puts forward a method of double valves compensation feedback control to depress the redundant force by analyzing the generation mechanism of the redundant force. Analysis and simulation results show that double valves compensation feedback control can be effective to compensate redundant force of load system, greatly improve the performance of robust and trace, and ideally depress redundant force of general load system.
Research on Selection and Interaction For set of Complex 3-D Entities in Simulation Model TONG Li-li, LIU Xiao-ping Among the commercial graphics packages of Computer Aided Design, there are some limitations about display and selection of complex 3D entities. In this paper, a new method which check relationship between point and region is proposed. It based on and improved on radial algorithm about scan-transition-fill in computer graphics. Paper study interactive techniques for complex 3-D multi-entities also. Experiment result at the end of the article verifies the effectiveness of presented approach.
Virtual Battlefield Simulation of Infrared Imaging Guiding Missile WANG Xue-wei, XIONG Zhang,

SHEN Tong-sheng, ZHOU Xiao-dong

The battlefield environment of the infrared imaging guiding missile includes the target, backgrounds and interference which are detected by the guider. A simulation method was introduced to generate the virtual battlefield which was generated by five interactive models: target, background, interference, sensor and atmosphere. The targets and interference were combined into the backgrounds by analyzing their infrared radiant characteristic using the CIG to generate the successive infrared images. It can achieve the arbitrary changing of the battlefield. The battlefield adaptability of the guiding system can be evaluated in laboratory after the guider simulation module is added in.
An MPEG-4 Compliant Speech Animation System WANG Xun, ZHANG Dao-yi, DONG Lan-fang, CHEN Guo-liang Based on the MPEG-4 compliant facial animation system we developed, an MPEG-4 compliant speech animation system was designed and implemented. The MPEG-4 is the first international standard at facial animation. In this paper, we give the definition of the standard and the structure of our MPEG-4 compliant speech animation system and then, describe in detail the key techniques in implementation, including facial animation, the definition of Chinese viseme, viseme transmission, expression transmission, viseme and expression blending, and speech-animation synchronization. Because the computing method is simple and the communication cost is small, the system can produce high-quality speech animation with expression in an ordinary microcomputer
Dynamic Scheduling of Batch Processes Based on Simulation of Petri Nets LI Hui-fang, FAN Yu-shun Batch processes usually exist in a dynamic and complex environment. The objective or subjective disturbance, as well as inherent process changeability, will lead to the changing of various process parameters, such as the perturbation of processing times, the changing of raw material quality, etc. These changes may make the existing static scheduling lose its optimality and even become unfeasible. So, it is required to consider dynamic scheduling with quick responsivity to unexpected events so as to catch the real time changes in production environment. For the deviations of processing times more often occurring in batch processes, a dynamic rescheduling method is put forward in this paper by integrating with rule-based schedule and Petri net simulation. This method can effectively catch the deviations of processing times and then improve the scheduling performance. It indicates a new direction for the research in dynamic scheduling of batch processes. Some simulation examples are demonstrated for this method.
ColuSimu: A Simulation Software for the Research of Model Reduction and Control of Distillation Columns WANG Wei-bo, LI Shu-rong ColuSimu is an integrated visual simulation software for the research of model reduction and control of distillation columns, which is developed by us on the basis of theoretical results in this area. It contains simulation modules for the research of model reduction, control structures, nonlinear control based on full-order model and nonlinear control based on reduced-order model. It is easy to perform various simulation studies on distillation columns by the software.
Chebyshev Fit of Propeller Properties across Four Quadrants and Realization of All-round Movement Simulation along X-shift of DSV LI Dian-pu, WANG Zong-yi, CHI Hai-hong A Chebyshev fit expression across four quadrants for a propeller in Deep Submergence Vehicle is given concretely and is changed to ordinary polynomial expression. The dynamic simulation models of propeller corresponding four quadrant work are given. Therefore, it becomes practical to express static and dynamic properties of shift load of the DSV propulsive motor. On this basis, the dynamic propulsive models of DSV across four quadrants are set. The simulations show the results are very close to the practical test data and the correctness and effectiveness of above results are proved.
Run Simulation of Self-mechanized Gun Using Adams Tracked Vehicle MA Ji-sheng, YAN Shi In this paper, a dynamic model of a self-mechanized gun is created by adding rotating gun turret, cradle assembly and recoil gun barrel to the tank model in ATV. There are 2 track systems in this model and each track system is constituted of 1 idler, 6 roadwheels, 1 sprocket, 3 support rolls and 84 segments. The whole model has 1051 freedom degrees. The forces acting on the idler, roadwheels, sprocket, support rolls and segments are discussed in detail. 4 simulations are implemented under the following conditions: (1) setting the same rotation speed for the 2 sprockets, letting the gun cross an obstacle.  (2) setting the same drive torque for the 2 sprockets, letting the gun run on a plain road. (3) setting drive torque for one sprocket and 0 for another sprocket, letting the gun turn on a plain road. (4) setting drive torque for the two sprockets with the same amount and different sign, letting the gun turn on a plain road. Based on the above analysis, the capability of the gun can be judged quantificationally.
Integrated Control Simulation for Power System with Superconducting Magnetic Energy Storage GE You, LI Chun-wen For single machine infinite bus power system with superconducting magnetic energy storage (SMES), this paper constructs differential algebraic equations (DAE) system model on the Simulink platform of Matlab. Combined inverse system method and small signal model of a SMES unit, SMES integrated controller is designed. The Simulink platform-based simulations under two conditions show that proposed controller can improve power system transient performance effectively.
An Approach to Multitask Scheduling for Underwater Vehicle Guidance Simulation System LIU Wei-dong, GAO Li-e, XU De-min,

SHENG Meng, KONG Feng

A new approach is introduced for simulation of underwater vehicles. The dynamic model, control and guidance equations of the underwater vehicles are discussed, and the P/T model for this multitask system is described in details. In order to make the multitask scheduling be realized by system management scheduler of the Operating System, modified solutions for join and fork models are presented. Finally, some simulation results are given to demonstrate the validity of the proposed methods.
Computer Simulation of A Complex Crank Rocker Arm Mechanism Moving Shao Lu-Shou, Jiao Jun Due to complex mechanical motion of the crank rocker arm of rice transplanter, we depend on miscellaneous math inference in teaching, which is insipid and indirect. With variation of parameters and motion rod, its motion is volatile. Through simulink, we establish mathematical modeling and carry out simulation, output graphs of motion locus,transplant rice seedlings' velocity and acceleration. At the same time we simulate results of mechanical optimum design and provide their graphs. These simulation results, videotape of static and dynamic state of crank rocker arm mechanism and practice in farmland have been used in multi-media teaching and receive good effect.
Design and Study of Observer for Linear Systems with Unknown Input YAN Wei-dong, LU Tian-sheng, YU-Yan For a class of linear systems with unknown input, this paper develops a new observer of state /unknown input and its existing condition. The strict verification to the observer is given. Through a example of linear system, using the proposed method, the observer of the linear system is designed. The simulation result shows that this kind of observer using only measured output, can asymptotically estimate the system’s state and unknown input to any desired accuracy.
Design of Discrete Fuzzy System Using New Fuzzy Controller and Observer WU Zhong-qiang, GAO Mei-jing Based on T-S fuzzy dynamitic model, the design analysis for discrete system using linear supervise controller in state space framework, has been studied. The global fuzzy model has been newly divided into some subspace by membership. Using subsection Lyapunov function method, the defect of solving a common positive matrix P in formerly is overcome and the Riccati equation in trouble needs not to be solved. The fuzzy observer is designed for system in which has immeasurable state variable in general. It is completed by using mature linear system theory to control complex nonlinear system. The fuzzy controller and observer are simple, and suit for engineering use.
A Low Probability of Intercept Direct Sequence Spread Spectrum Systems LUO Wei-bing, WANG Yong-sheng Some special detector, such as square detector and bit delay detector, can easily detect traditional Binary Phase-Shift-Keying (BPSK) modulation signal of direct sequence spread spectrum (DSSS) systems, which was a dormant disadvantage of carry-wave hiding. A new modulation can improve the performance of low probability of intercept (LPI) for DSSS systems by a multi-level pseudo-noise (PN) code. Performance of LPI in the new way has been simulated by SystemView software and compared the result with the traditional BPSK modulation. As result, the modulation can effectually avoid the detection by square detector and bit delay detector.
Fuzzy Control System Design of Multi-phase Traffic Signals in Isolated Intersection CHEN Shu-yan, CHEN Sen-fa, HUANG Yi It is not necessary for fuzzy control to built exact mathematic model about the controlled object, so fuzzy control is applicable especially to control urban traffic with great random characteristic. Having the problems of multi-phase and time backwards be considered, the traffic signal control employing fuzzy control method in isolated intersection is discussed, and a fuzzy control method is proposed which uses current phase and next phase vehicles waiting queue to determine each phase time. With the emphasis on design of the two-dimensional fuzzy controller, the fuzzy control system design scheme is given. Watch Dog in Mcs96 is exploited to ensure the system stability and to raise resisting interference capability of the system. Because each phase time is determined once according to the result of fuzzy inference about real time data of crossing, this system can be used at intersection with time backwards equipment, and it has favorable practicability.
A New Approach to the Sliding Mode Control Based on Fuzzy Reaching Law JIANG Kun, ZHANG Jing-gang A scheme of the sliding mode control based on fuzzy reaching law for a class of nonlinear systems with uncertainties is proposed in this paper. Since the proper fuzzy logic control is applied to the reaching law, the quality is improved. At the same time, the chatting phenomenon is avoided. Simulation results show that the control system exhibits not only strong robustness and better quality control but also avoiding the chatting phenomenon.
Improved Adaptive Generalized Predictive Poleplacement Weighted Controller CHEN Li-zhi, CAI Qing-sheng, FANG Jin, ZHANG Quan An important problem worthy of being noticed is how to decrease calculative quantity and ensure real-time problems while the application of predictive control. The generalized predictive poleplacement weighted controller of the literature has been improved in this aspect, but it is still required to solve closed loop pole equations on-line, and the calculative quantity is still large. In this paper, a new scheme is proposed, which uses recursive algorithm with small calculative quantity to identify parameters of above-said equations, and at the same time, the recursive algorithm about Diophantine equations of literatures is popularized. The application indicates the validity of the scheme.