Summary of the invention
The object of the present invention is to provide a kind of method for evaluating state of hydraulic machine based on multilayer fuzzy matching algorithm
The technical solution that realizes the object of the invention is: a kind of method for evaluating state of hydraulic machine based on multilayer fuzzy matching algorithm, be divided into three links: information comparison, information fusion and judgement conclusion, at first, the information comparison link is by the operating mode staging treating with hydropress, be that the hydropress duty is divided into: static, down fast, compacting, pressurize, pressure release, backhaul, hydraulic die cushion ejects, return, stamping-out, under each state, for all parameters are set up data characteristics table, by the time image data and compare the corresponding data mark sheet, calculate the recency mutually of collection result and corresponding data mark sheet analog value, the phase recency of all amounts is set up close kilsyth basalt;
Secondly, the information fusion link is divided into two-layer, ground floor is divided into two parts again, first is the information fusion of treatment of simulated amount, the kind of collection capacity as required, be respectively every kind of amount and provide weights, providing of these weights is relevant with hydropress history run situation, calculates analog quantity general status index in conjunction with close kilsyth basalt and all kinds of analog quantity information weights; Second portion is to handle digital information to merge, and its phase recency value non-zero is one; The second layer provides the analog quantity weight, provides the digital quantity weight, calculates hydropress overall operation index;
Once more, judge the conclusion link, should set up state of hydraulic machine system experts database earlier, the data processing in this storehouse is as follows: various analog quantitys under the various typicalnesses of hydropress and the tables of data under digital quantity and the health status carry out putting in storage after the data message comparison link, produce close kilsyth basalt after the data process information comparison link of gathering in real time, every information in this close kilsyth basalt and the state of hydraulic machine system experts database is carried out similarity analysis, find out the most similar state in this close kilsyth basalt and the experts database system, and point out its similarity numerical value, the numeric representation of final state of hydraulic machine conclusion is divided into three parts: hydropress overall operation index, the most similar state of hydropress and similarity.
The present invention compared with prior art, its remarkable advantage: (1) can effectively solve the nonlinear problem of height of state of hydraulic machine model; (2) be under different behaviours in service and the health status at hydropress, can finish the evaluation of state of hydraulic machine flexibly.
Below in conjunction with accompanying drawing the present invention is described in further detail.
Embodiment
In conjunction with the accompanying drawings, the present invention is based on the method for evaluating state of hydraulic machine of multilayer fuzzy matching algorithm, form according to state of hydraulic machine monitoring principle design, sensor is handled the numerical Evaluation that provides the current operation conditions of hydropress to informix by information comparison, information fusion and three links of judgement conclusion.Information comparison is responsible for providing real time data close kilsyth basalt, and information fusion is based on the close kilsyth basalt of real time data, for every kind of signal is given weights, and carries out COMPREHENSIVE CALCULATING and provides analog quantity overall target, digital quantity index, hydropress integral status index.Judge that conclusion provides the method for building up of state of hydraulic machine system experts database, and, find out the highest state of similarity, and provide the similarity value based on the close kilsyth basalt of real time data.
The present invention is a three-decker: information comparison, information fusion and judgement conclusion.At first, the information comparison link is pressed the operating mode staging treating with hydropress.The hydropress duty generally is divided into: static, fast down, compacting, pressurize, pressure release, backhaul, hydraulic die cushion eject, return, stamping-out etc.Under each state, all pressure, flow, oil temperature, ram speed, slide position, hydraulic die cushion speed, position, relay on-off information are different, and still this tittle is basic identical or have the continuous variation that can estimate under the same operating mode.Under each state, for all parameters are set up data characteristics table, by certain acquisition time image data and compare the corresponding data mark sheet, calculate the recency mutually of collection result and corresponding data mark sheet analog value, the phase recency of all amounts is set up close kilsyth basalt.
Secondly, the information fusion link is divided into two-layer, ground floor is divided into two parts again, the information fusion of first's treatment of simulated amount, the kind of collection capacity as required, be respectively every kind of amount and provide weights, providing of these weights is relevant with hydropress history run situation, calculates analog quantity general status index in conjunction with close kilsyth basalt and all kinds of analog quantity information weights.Second portion is handled digital information and is merged, and its phase recency value non-zero is one, so digital quantity general status index only may be zero or one.The second layer provides the analog quantity weight, provides the digital quantity weight, calculates hydropress overall operation index.
Once more, judge the conclusion link, this link at first should be set up state of hydraulic machine system experts database, and the data processing in this storehouse is as follows: various analog quantitys under the various typicalnesses of hydropress and the tables of data under digital quantity and the health status carry out putting in storage after the data message comparison link.Produce close kilsyth basalt after the data process information comparison link of gathering in real time, every information in this close kilsyth basalt and the state of hydraulic machine system experts database is carried out similarity analysis, find out the most similar state in this close kilsyth basalt and the experts database system, and point out its similarity numerical value.The numeric representation of final state of hydraulic machine conclusion is divided into three parts: hydropress overall operation index, the most similar state of hydropress and similarity.
Information comparison link of the present invention is monitored principle based on state of hydraulic machine, and promptly hydropress is following in working order, and in each moment, all health status information are can be pre-determined.Hydropress is pressed the operating mode staging treating.The hydropress duty generally is divided into: static, fast down, compacting, pressurize, pressure release, backhaul, hydraulic die cushion eject, return, stamping-out etc.Under each state, all pressure, flow, oil temperature, ram speed, slide position, hydraulic die cushion speed, position, relay on-off information are different, and still this tittle is basic identical or have the continuous variation that can estimate under the same operating mode.Under each state, for all parameters are set up data characteristics table, by certain acquisition time image data and compare the corresponding data mark sheet, calculate the recency mutually of collection result and corresponding data mark sheet analog value, the computing formula of phase recency is:
Phase recency=(signal actual value-signal theory value)/abs (signal actual value-signal theory value) * signal actual value/signal theory value,----------------------------------------------------(1)
Wherein abs () is an ABS function
With the force value is example, supposes actual pressure value 14.5MPa, and the theoretical pressure value is 15MPa, and then the phase recency is:
Pressure phase recency=(14.5-15)/abs (14.5-15) * 14.5/15=-0.97, it is near that negative sign characterizes negative.
The phase recency of all amounts is set up close kilsyth basalt.
Information fusion link of the present invention is divided into two-layer, ground floor is divided into two parts again, the information fusion of first's treatment of simulated amount, the kind of collection capacity as required, be respectively every kind of amount and provide weights, providing of these weights is relevant with hydropress history run situation, calculates analog quantity general status index in conjunction with close kilsyth basalt and all kinds of analog quantity information weights.Second portion is handled digital information and is merged, and its phase recency value non-zero is one, so digital quantity general status index only may be zero or one.The second layer provides the analog quantity weight, provides the digital quantity weight, calculates hydropress overall operation index.
The computing formula of analog quantity operation index is:
Ad=∑Ki×Si/N?--------------------------------------------------(2)
Wherein Ad represents analog quantity operation index, and Ki represents i kind signal phase recency, and Si represents the weights of i kind signal with respect to the analog quantity operation, and N represents the species number altogether of signal.
The overall operation formula of index is:
Hd=(Ad×Sa+Dd×Sd)/2?--------------------------------------------(3)
Wherein Sa represents the whole weight of analog quantity, and Dd represents digital quantity phase recency, and Sd represents the whole weight of digital quantity.
The present invention judges the conclusion link, this link at first should be set up state of hydraulic machine system experts database, the data processing in this storehouse is as follows: various analog quantitys under the various typicalnesses of hydropress and the tables of data under digital quantity and the health status carry out putting in storage after the data message comparison link, data loading is handled formula homophase recency formula, and as following table: hydropress is judged the conclusion table.
Produce close kilsyth basalt after the data process information comparison link of gathering in real time, every information in this close kilsyth basalt and the state of hydraulic machine system experts database is carried out similarity analysis, find out the most similar state in this close kilsyth basalt and the experts database system, and point out its similarity numerical value.
Calculating formula of similarity:
M=∑abs(Wi)×Li/N?------------------------------------------------(4)
Wherein M represents analog quantity operation index, and Wi represents i kind signal phase recency, and Li represents the similarity weights of i kind signal, and N represents the species number altogether of signal, abs () ABS function.
The numeric representation of final state of hydraulic machine conclusion is divided into five parts: hydropress overall operation index Hd, hydropress analog quantity operation Index A d, digital quantity operation index D d, the most similar state of hydropress and similarity M.
Hydropress is judged the conclusion table
?M |
Hd |
Dd |
Ad |
Similar state |
|
|
|
|
|