| Coal mine is one of high dangerous profession, so safety in production is the always important. The water storage of coal mine is a important facility to prevent the flood of the pit and make insure the safe in producing. To solve the problem of cleaning of water storage and cleaning with mechanization, we have developed a kind of QCL03 bunker cleaner. The working set of the bunker clearer is RSCTM, and traditional optimization design methods is to adopt complex method and penalty function et al. that can not solve the multiple goal satisfactory optimization problems so this paper applies the method combined by optimization theory of satisfaction and ADAMS virtual prototype to do optimization design researches. 1. Establishment of mathematics model Select mechanism as the move reference system of the working set of RSCTM to establish XOY coordinate system. The analysis model of the working set of RSCTM is shown in Figure 1. 1.1 Design variable Set the following variable as the optimization design variable of the working set of RSCTM, it is that [ ]X = R2 ,R4,R5,R6,R8,UG1,UG2,UD1,UD3,UA1T 1.2 Goal function To guarantee the good dynamic behavior of the bunker cleaner, the biggest value of digging force should form in the location of digging on the ground. On the condition of satisfying the demand of dig force, it can make the working set optimization to improve the performance of move of the bucket. Therefore, the goal function of optimal design is to make the digging force of the bucket ram largest, at the same time, the stability of move of the bucket should be better. It is that ( )min1 2,4,5,6,8,1,2,1,3,11F = fFRRRRRUGUGUDUDUA ( )min ddUUG = fu R2,R4,R5,R6,R8,UG1,UG2,UD1,UD3,UA1 Where U is the position angle of the bucket, UG the position angle of the movable arm, F1 thrust of the bucket ram. 1.3 Constraint condition Upper limits and lower limits of ten design variables are set for linkage, twenty edge restraint conditions in aggregate, and the concrete data are shown in Table 1. Table 1 design variables variable R2 R4 R5 R6 R8 UG1 UG2 UD1 UD3 UA1 upper limit 170 430 720 290 130 60 15 160 60 80 lower limit 250 490 800 350 180 85 25 180 75 95 Generally constraints of performance include the following: ①Geometric constraint The harmony of linkages when moving should be satisfied. Interference and the position of dead center of linkages should be avoided while working set is in operation, this is the necessary condition on that the mechanism forms.②Requirement of the performance of move Besides those mentioned above, the performance of move still needs extra constraints UG=UG7, then γp62° UG=UG4, then 40°≤γ≤45°③Constraint of the limit stop When the movable arms move from the position of lower limits to the position of transportation, the bucket should keep contact with the limit stops of the movable arms, it is that the bar AB leans on the bar AD. ④Constraint of discharge It is the variable characteristic of the length of discharge of the bucket ram when the angle of discharge is 45 °. ⑤Constraint of set level automatically After the bucket discharges on the certain height(usually it is the altitude of Upper limit of discharge), the bucket has the characteristic of restoring the initial state of digging while the movable arms fall down to the position of lower limit. ⑥Constraint of angle of drive Angle of drive is generally the angle made with the positive bar by the passive bar in the course of lifting and discharge. Constraint of angle of drive is achieved by establishing constraint conditions of limit angle of drive, it is that 10 °≤UB ≤170°⑦Stability constraint of oil cylinder Generally it is required that the extension proportion of oil cylinder is less than 1.6, it is that 1.671RR 73 p ⑧Constraint of overall layout The constraint of overall layout is defined as the constraint established by the overall layout restriction. For QCL03, it is required that Distance of unload H 1≥1200 Altitude of unload H 2 ≥1100 ⑨Constraints else There are a few constraints for the design of the bunker cleaner, and where some important factors will be mentioned with the neglect of factors else.2 The satisfaction optimization of the working set of the QCL03 bunker cleaner The satisfaction optimization is the products combined by principle of the satisfaction degree and theory of optimization, and its opinion is that in solving the optimization, it searches the satisfaction solution, based on the practice, instead of the optimization one. 2.1 The basic concept of the satisfaction optimization of multiple objective The general forms of multiple objective problems are as follows: Where = is called vector goal function, VminF( x)? x∈Rnthe abbreviation for the minimal model of multiple objective (vector form), V-min minimal of vector, and the goal function can be minimized equally or not, g( x) h( x)和constraint condition. It is very difficult for the problem of multiple goal optimization that every goal gets optimum, especially there are contradiction between each goal, it is that there are conflicts among the solutions of each goals, and usually the optimum does not exist. (1) Goal satisfaction degree In a optimization problem there are one or more goal functions and we call the degree of one of feasible solutions of certain goal functions satisfying the goal satisfaction degree of the solution for the goal. (2) Constraint satisfaction degree In a constraint optimization problem, constraint condition must be satisfied, and the degree of a feasible solution accords to a certain constraint condition is called the constraint satisfaction degree of the constraint.(3) The problem of satisfaction degree of multiple goal optimization It is the degree of the solution satisfies both multiple goals and multiple constraint, written as S ( F,f ). 2.2 The solution of satisfaction optimism of the working set of the bunker cleaner The problem of optimization of the working set of the bunker cleaner belongs to the problem of multiple goal satisfaction optimizations with the difficulty of expression of satisfaction function, and it is also difficult to explain in analytic expressions or other relations, but satisfaction degree still exists probably with a certain principle. Through establishing a kind of multiplayer BP NN, and making use of the ability of corresponding satisfaction degree can be attained, a kind of hidden corresponding, and at the same time the research of the solution in the solution space also can be achieved by a certain algorithm. Then by relying on the BP network corresponding to the satisfaction degree of the solution, eventuallyappraise and the selection of the solution can be achieved. The overall diagram of the cource of the solution is shown in Figure 2. 2.3 The expression of satisfaction based on BP NN The NN is to establish the function relation of satisfaction degree of the problem, and is a structure with multiple inputs and one output, the output is the satisfaction degree of the solution, and the input is the value of design variable X1 of RSCTM. input layer(solution) hidden layer output layer(satisfaction degree) Based on the mathematical model of the working set of the bunker cleaner, We know that the design variable of the working condition X1 is 10, and the number of selected to input the layer of nerve cell amounts to the dimension of input vector (10). On the consideration of the NN including an intricate relation of corresponding, the number of the hidden layer units is large, m = 15. The solution of output layer is the appraisal satisfaction degree. Therefore the structure of NN of the design is one 10-15-1 three-layer BP network, shown in Figure 3. 2.4 The training of network Through the arrange of the structure parameters of the working set of the bunker cleaner put into practice, the train data of the network has been attained, shown in Table 2. Table 2 is the appraisal results of the design (satisfaction degree) attained by arraying the sample of data of the two-design variable of the working set of the QCL bunker cleaner and some investigates. 2.5 Solution of simulation optimization based on ADAMS virtual prototype Establish the model of virtual prototype of the working set of the bunker cleaner with the design parameter of original scheme, and carry out kinematics simulation, shown in Figure 4. FIG.4 model of prototype of RSCTM FIG.5 Optimization results of optimal design The optimization design scheme based on ADAMS from figure 5 is shown in Table 3. Table 3 The variable value after optimization R2 R4 R5 R6 R8 UG1 UG2 UD1 UD3 UA1 190 459 743 306 152 65 20 170 66 84 Table 4 Satisfaction optimization solution of the working set of the bunker cleaner R2 R4 R5 R6 R8 UG1 UG2 UD1 UD3 UA1 Satisfaction degree 190 459 743 306 152 65 20 170 66 84 0.82... |
PDF Full Text Request