JP2023108526A - Characteristic value estimation device and characteristic value estimation method - Google Patents

Abstract

To provide a characteristic value estimation device and a characteristic value estimation method capable of easily estimating a characteristic value of a control object.SOLUTION: A characteristic value estimation device 1 estimates a characteristic value of a control object W. The characteristic value estimation device 1 includes: a first characteristic value acquisition unit 11 which determines a first characteristic value of the control object W from a frequency characteristic of the control object W; a second characteristic value calculation unit 13 which calculates a second characteristic value of the control object W by using a state equation containing a variable which influences the first characteristic value, and changing the variable in a search region; an optimization processing unit 14 which obtains an optimum solution, in the search region, of an evaluation function being a relational expression of the first characteristic value and the second characteristic value by changing the second characteristic value calculated by the second characteristic value calculation unit 13 while changing the variable in the search region; and an estimation unit 15 which estimates the variable when the optimum solution is obtained by the optimization processing unit 14 as the characteristic value.SELECTED DRAWING: Figure 1

Description

The present invention relates to a characteristic value estimation device and a characteristic value estimation method for estimating characteristic values of a controlled object.

Motors (dynamos, dynamometers) apply loads to specimens (drive motors, engines, transmissions, and other power systems) in automobile testing equipment that performs evaluation tests on specimens. In a power transmission system in which a torque detector or the like is axially coupled, torsional vibration generated in each coupling shaft is a problem. As one of the solutions, it is considered to apply a disturbance observer or the like to the automobile test equipment.

Also, as a method for seeking solutions to problems of controlled objects, a method of analyzing controlled objects using simulation is known. In the case of such a method, it is necessary to reflect the characteristic values of the controlled object in the simulation.

In any of the above cases, it is necessary to estimate the characteristic value of the controlled object. In this way, in the field of control, there is a demand for estimating characteristic values of controlled objects. As a method for estimating the characteristic value of such a controlled object, for example, a parameter identification device disclosed in Patent Document 1 is known.

The parameter identification device of Patent Document 1 identifies model parameters of a two-inertia system model. Specifically, the parameter identification device generates a plurality of phase plane diagrams based on a torque command for a motor, measured values of the angle and angular velocity of the motor, and measured values of the angle and angular velocity of the load, and Model parameters representing the inertia, viscosity, and friction of the motor, the inertia, viscosity, and friction of the load, the rigidity of the connecting member that connects the motor and the load, and the width of the dead band of the connecting member, based on the plan view. identify.

Moreover, as a method of estimating the characteristic value of the controlled object, there is also a method of estimating the characteristic value by obtaining a numerical formula for fitting to the time-series data using a calculation program.

Japanese Patent No. 6756653

By the way, the parameter identification device of Patent Document 1 is intended for a two-inertia system model, and it is necessary to actually measure the angular velocities of the motor and the load. Therefore, if the parameter identification device of Patent Document 1 is used to estimate the characteristic values of the controlled object even in the case of a multi-inertia system, it is necessary to significantly change the algorithm of the parameter identification device. A large number of sensors are required to actually measure the angular velocity.

In addition, the method of estimating characteristic values by finding mathematical formulas that fit time series data also requires time for calculation, and it is difficult to easily estimate characteristic values that can accurately fit mathematical formulas to time series data. is difficult.

An object of the present invention is to provide a characteristic value estimating device and a characteristic value estimating method that can easily estimate characteristic values of a controlled object.

A characteristic value estimation device according to one embodiment of the present invention is a device that estimates a characteristic value of a controlled object. This characteristic value estimating apparatus includes a first characteristic value acquisition unit that obtains a first characteristic value of the controlled object at a predetermined frequency from the frequency characteristic of the controlled object, and a state that includes a variable that affects the first characteristic value. a second feature value calculator that calculates a second feature value of the controlled object when the variable is changed within the search area at the predetermined frequency using an equation; and the variable is changed within the search area. By changing the second feature value calculated by the second feature value calculating unit, the evaluation function, which is the relational expression between the first feature value and the second feature value, is optimized in the search area. It has an optimization processing unit that obtains a solution, and an estimation unit that estimates the variable when the optimum solution is obtained by the optimization processing unit as the characteristic value (first configuration).

Thereby, the variables of the state equation for obtaining the characteristic value of the controlled object are changed within the search area to obtain the optimum solution of the evaluation function, and the variable when the optimum solution is obtained is estimated as the characteristic value of the controlled object. be able to. The evaluation function is a relational expression between a first feature value obtained from the frequency characteristic of the controlled object and a second feature value calculated using the variable and the state equation changed within the search area. Therefore, the characteristic value can be easily estimated by obtaining the optimum solution of the evaluation function.

By the way, in the case of time response fitting and frequency response fitting, which are generally performed, it is necessary to fit data to each measured value. Therefore, the fitting accuracy is affected by the amount of data. Therefore, in the fitting of the time response and the fitting of the frequency response, it is necessary to fit data to many measurement points in order to perform fitting with high accuracy, and the computational load of the apparatus is large.

On the other hand, in the above-described configuration, optimization is performed by the evaluation function using the first feature value and the second feature value, so the data arithmetic processing can be greatly reduced.

Therefore, it is possible to reduce the arithmetic processing of the device compared to the conventional case of obtaining characteristic values to be fitted to time-series data or the like.

Therefore, with the above configuration, a characteristic value estimating device that can easily estimate the characteristic value of the controlled object is obtained.

In the first configuration, the evaluation function is a relational expression including the square of the difference between the first feature value and the second feature value. The optimization processing unit obtains the value of the evaluation function when the square of the difference between the first feature value and the second feature value is the smallest as the optimum solution (second configuration).

Thus, in the first configuration, it is possible to realize an evaluation function that is a relational expression between the first feature value and the second feature value. Moreover, by squaring the difference between the first feature value and the second feature value, the difference between the first feature value and the second feature value can be obtained regardless of the magnitude of the first feature value and the second feature value. The optimization processing unit sets the value of the evaluation function when the square of the difference between the first feature value and the second feature value is the minimum as the optimum solution of the evaluation function, so that the optimum solution can be easily obtained. can be done.

Therefore, with the above configuration, a characteristic value estimating device that can easily estimate the characteristic value of the controlled object is obtained.

In the first or second configuration, the first feature value and the second feature value each include at least a gain (third configuration).

Thereby, using the vibration frequency and the gain, the characteristic value such as the spring constant of the controlled object can be obtained.

In any one of the first to third configurations, the search area is composed of a plurality of local areas. The optimization processing unit changes the second feature value calculated by the second feature value calculation unit by changing the variables of the state equation in each of the plurality of local regions, a local region optimization processing unit that obtains an optimal solution of an evaluation function; and an optimal solution among the optimal solutions of the evaluation function obtained in each local region by the local region optimization processing unit, the evaluation function in the search region. (fourth configuration).

As a result, the optimum solution of the evaluation function can be obtained in each of the local regions obtained by dividing the variable search region into a plurality of regions. Then, by obtaining the optimum solution in the search area from among the optimum solutions in each local area, it is possible to obtain the optimum solution for the entire search area instead of the local optimum solution. Therefore, the optimum solution of the evaluation function in the search area can be obtained with high accuracy. Therefore, the characteristic value of the controlled object can be obtained with high accuracy.

Moreover, with the above configuration, the characteristic value can be estimated even if the search area is wide, so the characteristic value can be obtained with high accuracy without adjusting the range of the search area and without obtaining a local solution. .

On the other hand, if the variable search area is not divided into a plurality of areas, there is a high possibility that a local solution will be found, so it is necessary to change the search area through trial and error. Also, although it is conceivable to avoid the calculation of the local optimum by increasing the number of iterations of the optimization or by tightening the judgment value of the convergence evaluation of the evaluation function, the calculation takes time. On the other hand, by dividing the search area into a plurality of local areas as in the configuration described above, it is possible to obtain the optimum solution without trial and error or changing the judgment value for convergence evaluation. Therefore, by dividing the search area into a plurality of areas as described above and finding the optimum solution of the evaluation function in each local area, the calculation load of the characteristic value estimation device can be reduced.

A characteristic value estimation method according to an embodiment of the present invention is a method of estimating a characteristic value of a controlled object. This characteristic value estimating method includes a first characteristic value acquiring step of obtaining a first characteristic value of the controlled object at a predetermined frequency from the frequency characteristic of the controlled object; Using an equation, a second characteristic value of the controlled object is calculated when the variable is changed within the search area at the predetermined frequency, and a relational expression between the first characteristic value and the second characteristic value is obtained. and an estimation step of estimating the variable when the optimum solution is found in the optimization process as the characteristic value ( first method).

Thus, the variables of the state equation for obtaining the characteristic value of the controlled object are changed within the search area to obtain the optimum solution of the evaluation function, and the variable when the optimum solution is obtained is estimated as the characteristic value of the controlled object. be able to. The evaluation function is a relational expression between a first feature value obtained from the frequency characteristic of the controlled object and a second feature value calculated using the variable and the state equation changed within the search area. Therefore, the characteristic value can be easily estimated by obtaining the optimum solution of the evaluation function.

Therefore, it is possible to realize a characteristic value estimation method that can easily estimate the characteristic value of the controlled object.

A characteristic value estimating device and a characteristic value estimating method according to an embodiment of the present invention obtain a first characteristic value of the controlled object at a predetermined frequency from the frequency characteristic of the controlled object, and determine the first characteristic value at the predetermined frequency. The second feature value is changed by changing the variable influencing the value within the search area, and the optimum solution in the search area of the evaluation function, which is the relational expression between the first feature value and the second feature value. Ask for Then, the characteristic value estimating device and the characteristic value estimating method estimate the variable when the optimum solution is obtained as the characteristic value.

Thereby, a characteristic value estimation device and a characteristic value estimation method that can easily estimate the characteristic value of the controlled object can be obtained.

FIG. 1 is a diagram showing a schematic configuration of a characteristic value estimating device according to Embodiment 1 using functional blocks. FIG. 2 is a diagram showing an example of a controlled object as a model. FIG. 3 is a Bode diagram showing an example of frequency characteristics of a controlled object represented by a state equation. FIG. 4 is a diagram showing the flow of the characteristic value estimation method. FIG. 5 is a diagram showing an example of a search area. FIG. 6 is a diagram showing a schematic configuration of a characteristic value estimating device according to the second embodiment using functional blocks. FIG. 7 is a diagram showing the flow of the characteristic value estimation method according to the second embodiment. FIG. 8 is a diagram showing an example of local regions obtained by dividing the search region. FIG. 9 is a diagram showing an example of frequency characteristics obtained using variables when the local solution of the evaluation function is obtained. FIG. 10 is a diagram showing an example of frequency characteristics when the optimum solution of the evaluation function is obtained by the method of the second embodiment. FIG. 11 is a diagram showing, in functional blocks, a schematic configuration of a characteristic value estimation device according to the third embodiment. FIG. 12 is a diagram schematically showing the relationship between values used in the evaluation function. FIG. 13 shows an example of frequency characteristics when the moment of inertia, spring constant, and viscous damping coefficient differ between the measured values and the calculated values. FIG. 14 shows an example of frequency characteristics when the viscous damping coefficient differs between the measured value and the calculated value. FIG. 15 shows an example of frequency characteristics when the spring constant differs between the measured value and the calculated value.

BEST MODE FOR CARRYING OUT THE INVENTION Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings. The same reference numerals are given to the same or corresponding parts in the drawings, and the description thereof will not be repeated.

<Embodiment 1>
FIG. 1 is a diagram showing a schematic configuration of a characteristic value estimation device 1 according to Embodiment 1 of the present invention using functional blocks. This characteristic value estimating device 1 is a device for estimating characteristic values of a controlled object W, and is configured by an arithmetic device such as a computer, for example.

(controlled object)
The controlled object W is, for example, a multi-inertia system. FIG. 2 is a diagram showing an example of the controlled object W as a dynamic characteristic model of a mechanical system. A controlled object W shown in FIG. 2 is a four-inertia system. The controlled object W includes a motor, a torque detector, a test piece, and the like. In the example shown in FIG. 2, j1 is the moment of inertia of the motor, and j2 to j4 are the moments of inertia of the torque detector, test piece, and the like.

As shown in FIG. 2, structures with moments of inertia j1-j4 are connected by springs with spring constants k1-k3 and dampers with viscous damping coefficients c1-c3. A command torque u is input to a motor having a moment of inertia j1. Note that the torsional torque generated between structures in the controlled object W is T _i =k _i ∫(ω _i −ω _i+1 )dt. In the example shown in FIG. 2, i=1,2,3.

In the controlled object W, when the torsional torque of the shaft between the structure having the moment of inertia j2 and the structure having the moment of inertia j3 can be detected by a sensor, an example of the state equation is as follows. Note that the state equation may be other than the following equations.

FIG. 3 is a Bode diagram showing an example of the frequency characteristics of the controlled object W represented by the state equation described above. As shown in FIG. 3, the controlled object W, which is a four-inertia system, generally has three resonances.

(Characteristic value estimation device)
As shown in FIG. 1, the characteristic value estimation device 1 includes a first characteristic value acquisition unit 11, a parameter setting unit 12, a second characteristic value calculation unit 13, an optimization processing unit 14, and an estimation unit 15. Prepare.

The first characteristic value acquisition unit 11 obtains the resonance frequency and its gain from the measurement results of the input/output frequency characteristics of the object W to be controlled. Specifically, the first feature value acquisition unit 11 acquires three resonance frequencies and their gains in the Bode diagram shown in FIG. The first characteristic value acquiring unit 11 may automatically acquire the resonance frequency and gain of the controlled object W, or the resonant frequency and gain of the controlled object W are input to the first characteristic value acquiring unit 11. may The first characteristic value acquisition unit 11 may have any configuration as long as it can acquire the resonance frequency of the controlled object W and its gain.

In the present embodiment, the resonance frequency and its gain obtained from the frequency characteristic of the controlled object W by the first characteristic value acquisition unit 11 correspond to the first characteristic value. In the present embodiment, the first characteristic value is the eigenvalue of the frequency characteristic of the object W to be controlled. The first characteristic value may include parameters other than the resonance frequency and its gain as long as the first characteristic value is the characteristic value of the object W to be controlled. That is, the first feature value may be a value other than the eigenvalue. In this embodiment, the resonance frequency is the predetermined frequency in the present invention.

The parameter setting unit 12 acquires, for example, the inertia moments j1 to j4 and the spring constant k2 from the design information and frequency characteristics of the controlled object W. The parameter setting unit 12 may automatically acquire the parameters of the controlled object W, or the parameters of the controlled object W may be input to the parameter setting unit 12 . The parameter setting unit 12 may have any configuration as long as it can acquire the parameters of the controlled object W. FIG.

The parameter setting unit 12 also sets a search area for variables that are the remaining parameters (eg, spring constants k1 and k3 and viscous damping coefficients c1 to c3) in the state equation of the controlled object W. FIG. Specifically, the parameter setting unit 12 sets a search area for the variables to be calculated by the optimization processing unit 14, which will be described later. The search area for the variable may be stored in advance in a storage unit (not shown) or the like, or may be input to the parameter setting unit 12 .

The second feature value calculator 13 uses the state equation of the controlled object W to obtain the resonance frequency and its gain after changing the variables of the state equation within the search area. The second feature value calculator 13 obtains the resonance frequency and the gain using the state equation when the variable is changed within the search area by the optimization processor 14, which will be described later. That is, the second feature value calculation unit 13 obtains the resonance frequency and the gain using the variables changed in the search area when the optimization processing unit 14 performs the calculation of the optimization processing.

In the present embodiment, the resonance frequency and its gain calculated by the second feature value calculator 13 using the changed variables correspond to the second feature value. In the present embodiment, the second feature value is the eigenvalue of the frequency characteristic of the object W to be controlled. The second characteristic value may include parameters other than the resonance frequency and its gain as long as it is the characteristic value of the controlled object W. That is, the second feature value may include values other than the eigenvalue. In this embodiment, the resonance frequency is the predetermined frequency in the present invention.

The optimization processing unit 14 changes the variables within the search area, and uses the changed variables to obtain the second feature values obtained by the second feature value calculation unit 13 and the first feature value acquisition unit 11 to obtain Using the obtained first feature value, an optimization process is performed to obtain an optimum solution of the evaluation function within the search area. Specifically, the optimization processing unit 14 uses the resonance frequency and its gain, which are the first feature values, and the resonance frequency and its gain, which are the second feature values, to determine whether the evaluation function is the minimum in the search area. Find the optimal solution that becomes , and find the variables at that time.

The evaluation function is a relational expression between the first feature value and the second feature value. That is, the evaluation function is a relational expression including the square of the difference between the first feature value and the second feature value. Specifically, the evaluation function is the square of the difference between the resonance frequency that is the first feature value and the resonance frequency that is the second feature value, the gain that is the first feature value, and the second feature value. It is a relational expression including the square of the difference with the gain. In the present embodiment, the evaluation function is a feature value of a plurality of modes (three modes of primary resonance, secondary resonance, and tertiary resonance in this embodiment) in the controlled object W, and a weighting factor for each mode. It is a relational expression added in consideration. Specifically, the evaluation function is, for example, the following formula (1).

Here, w1 to w6 are weighting coefficients, f1 to f3 are resonance frequencies that are first feature values, G(f1) to G(f3) are gains that are first feature values, and fm1 to fm3 is the resonance frequency, which is the second feature value, and G(fm1) to G(fm3) are the gains, which are the second feature values.

The optimization processing unit 14 can obtain a second feature value close to the first feature value by obtaining a value that minimizes the evaluation function when the variable is changed within the search range.

The estimation unit 15 estimates the variable at that time as the characteristic value of the controlled object W when the optimization processing unit 14 obtains the minimum value of the evaluation function.

(Characteristic value estimation method)
Next, a method for estimating the characteristic value of the controlled object W using the characteristic value estimating device 1 having the above configuration will be described with reference to FIG. FIG. 4 is a diagram showing the flow of the characteristic value estimation method.

When the flow shown in FIG. 4 starts (START), first, the first characteristic value acquisition unit 11 acquires the frequency characteristic of the controlled object W in step SA1. Then, at step SA2, the first feature value acquisition unit 11 obtains the resonance frequency and its gain from the frequency characteristics. That is, the first feature value acquisition unit 11 acquires the first feature value.

In the subsequent step SA3, the parameter setting unit 12 acquires the inertia moments j1 to j4 and the spring constant k2 from the design information, frequency characteristics, etc. of the controlled object. That is, the parameter setting unit 12 acquires constants other than variables. Thereafter, at step SA4, the parameter setting unit 12 determines the search area for the variables. FIG. 5 is a diagram showing an example of a search area. For the sake of explanation, FIG. 5 shows a search area when the spring constants k1 and k3 are changed as an example of the variable search area.

At step SA5, the optimization processing unit 14 changes the variables within the search area. The optimization processing unit 14, for example, sets the initial value of the variable to the minimum value within the search area. The second characteristic value calculator 13 uses the changed variable and the state equation of the controlled object W to obtain the second characteristic value. The optimization processing unit 14 uses the first feature value acquired by the first feature value acquisition unit 11 and the second feature value acquired by the second feature value calculation unit 13 to obtain an optimal solution of the evaluation function. conversion processing. Specifically, the optimization processing unit 14 uses the resonance frequency and its gain, which are the first feature values, and the resonance frequency and its gain, which are the second feature values, to find the optimum solution that minimizes the evaluation function. and the variables at that time.

At step SA6, the estimating unit 15 estimates the variable at that time as the characteristic value of the controlled object W when the optimization processing unit 14 obtains the optimum solution of the evaluation function. After that, this flow ends (END).

In the above-described flow, step SA2 corresponds to the first feature value acquisition process, step SA5 corresponds to the optimization process, and step SA6 corresponds to the estimation process.

As described above, the characteristic value estimation device 1 of the present embodiment estimates the characteristic value of the controlled object W. FIG. The characteristic value estimation device 1 uses a first characteristic value obtaining unit 11 for obtaining a first characteristic value of the controlled object W from the frequency characteristic of the controlled object W, and a state equation including variables affecting the first characteristic value. a second feature value calculator 13 for calculating a second feature value of the controlled object W when the variable is changed within the search area; An optimization processing unit that changes the second feature value calculated by the value calculation unit 13 and obtains an optimum solution in the search region of the evaluation function, which is a relational expression between the first feature value and the second feature value. 14, and an estimation unit 15 for estimating the variable when the optimum solution is obtained by the optimization processing unit 14 as the characteristic value.

As a result, the variables of the state equation for obtaining the characteristic values of the controlled object W are changed within the search area to obtain the optimum solution of the evaluation function, and the variables when the optimum solution is obtained are estimated as the characteristic values of the controlled object W. can do. The evaluation function is a relational expression between a first feature value obtained from the frequency characteristic of the controlled object W and a second feature value calculated using the variable and the state equation changed within the search area. Therefore, the characteristic value can be easily estimated by obtaining the optimum solution of the evaluation function.

Therefore, the characteristic value estimating device 1 capable of easily estimating the characteristic value of the controlled object W is obtained with the above configuration.

Further, in this embodiment, the evaluation function is a relational expression including the square of the difference between the first feature value and the second feature value. The optimization processing unit 14 obtains the value of the evaluation function when the square of the difference between the first feature value and the second feature value is the smallest as the optimum solution.

By squaring the difference between the first feature value and the second feature value, the difference between the first feature value and the second feature value can be obtained regardless of the magnitude of the first feature value and the second feature value. The optimization processing unit 14 easily obtains the optimum solution because the value of the evaluation function when the square of the difference between the first feature value and the second feature value is the smallest is the optimum solution of the evaluation function. be able to.

Therefore, the characteristic value estimating device 1 capable of easily estimating the characteristic value of the controlled object W is obtained with the above configuration.

Further, in the present embodiment, the first feature value and the second feature value respectively include vibration frequency and gain. Thereby, the characteristic value such as the spring constant of the controlled object W can be obtained using the vibration frequency and the gain.

<Embodiment 2>
FIG. 6 is a diagram showing a schematic configuration of the characteristic value estimation device 100 according to the second embodiment using functional blocks. The characteristic value estimating device 100 is different in configuration from the characteristic value estimating device 1 of the first embodiment in that the variable search region is divided into a plurality of regions. In the following description, the same reference numerals are given to the same configurations as in the first embodiment, the description thereof is omitted, and only the configurations different from the first embodiment will be described.

(Characteristic value estimation device)
As shown in FIG. 6, the characteristic value estimation device 100 includes a first characteristic value acquisition unit 11, a parameter setting unit 120, a second characteristic value calculation unit 13, an optimization processing unit 140, and an estimation unit 15. have.

Note that in this embodiment as well, the first feature value is the resonance frequency and its gain obtained from the frequency characteristics of the object W to be controlled. The first feature value may be an eigenvalue or a value other than the eigenvalue. Also, the second characteristic value is the resonance frequency and its gain obtained from the frequency characteristic of the object W to be controlled. The second feature value may be an eigenvalue or a value other than the eigenvalue. The resonance frequencies in the first feature value and the second feature value are predetermined frequencies in the present invention.

The parameter setting section 120 has a search area dividing section 121 . The search area division unit 121 obtains a plurality of local areas by dividing the variable search area into a plurality of areas. The search area dividing unit 121 may divide the search area by a preset number, or may divide the search area into a plurality of areas according to a set variable range. The search area dividing unit 121 may divide the search area into a plurality of local areas according to the input such as the number of divisions and the range of division.

Similar to the parameter setting unit 12 of the first embodiment, the parameter setting unit 120 acquires, for example, the moments of inertia j1 to j4 and the spring constant k2 from the design information and frequency characteristics of the controlled object W.

The optimization processing unit 140 changes the variables in each local region obtained by the parameter setting unit 120, and uses the changed variables in each local region to calculate the second feature value calculation unit 13. Using the two feature values and the first feature value acquired by the first feature value acquisition unit 11, optimization processing is performed to find the optimum solution of the evaluation function. That is, the optimization processing unit 140 obtains the optimum solution of the evaluation function in each local region. The optimization processing unit 140 obtains the optimum solution in the search area from among the optimum solutions of the evaluation function obtained in the various areas.

Specifically, the optimization processing unit 140 has a local area optimization processing unit 141 and an optimal solution selection unit 142 .

The local region optimization processing unit 141 optimizes the evaluation function in each local region using the second feature value obtained by changing variables in each local region and the first feature value. find the solution. The local region optimization processing unit 141 may sequentially obtain the optimum solution of the evaluation function within a plurality of local regions, or may simultaneously obtain the optimum solution of the evaluation function in at least some of the plurality of local regions. may be asked for. The optimum solution is the minimum value of the evaluation function in each local region.

Note that the evaluation function is the same as the evaluation function in the first embodiment. Also, the method of obtaining the optimum solution of the evaluation function in each of the local regions is the same as the method of obtaining the optimum solution of the evaluation function within the search region in the first embodiment. Therefore, a detailed description of the method of obtaining the optimum solution within each local region will be omitted.

The optimum solution selection unit 142 selects the optimum solution in the search area from among the optimum solutions of the evaluation function in each local area obtained by the local area optimization processing unit 141 . In this embodiment, the optimal solution selection unit 142 selects the smallest value among the optimal solutions of the evaluation function in each local region as the optimal solution in the search region. The optimum solution selection unit 142 outputs variables for obtaining the optimum solution in the search area to the estimation unit 15 . The estimation unit 15 estimates the variable as the characteristic value of the object W to be controlled.

(Characteristic value estimation method)
Next, a method of estimating the characteristic value of the controlled object W using the characteristic value estimating device 100 having the above configuration will be described with reference to FIG. FIG. 7 is a diagram showing the flow of the characteristic value estimation method according to the second embodiment.

When the flow shown in FIG. 7 starts (START), first, the first characteristic value acquisition unit 11 acquires the frequency characteristic of the controlled object W in step SB1. Then, at step SB2, the first characteristic value acquisition unit 11 obtains the resonance frequency and its gain from the frequency characteristics. That is, the first feature value acquisition unit 11 acquires the first feature value.

In the subsequent step SB3, the parameter setting unit 120 acquires the inertia moments j1 to j4 and the spring constant k2 from the design information, frequency characteristics, etc. of the controlled object. That is, the parameter setting unit 120 acquires constants other than variables. After that, at step SB4, the parameter setting unit 120 determines the search area of the variable. Steps SB1 to SB4 are the same as steps SA1 to SA4 of the flow shown in FIG. 4 of the first embodiment.

At step SB5, the search area dividing unit 121 of the parameter setting unit 120 divides the search area into a plurality of local areas. FIG. 8 is a diagram showing an example of local regions obtained by dividing the search region. For the sake of explanation, FIG. 8 shows the search area when the spring constants k1 and k3 are changed as an example of the search area.

At step SB6, the local region optimization processing unit 141 of the optimization processing unit 140 changes variables within each local region. The local region optimization processing unit 141, for example, sets the initial value of the variable to the minimum value within each local region. The second characteristic value calculator 13 uses the changed variable and the state equation of the controlled object W to obtain the second characteristic value. The local region optimization processing unit 141 uses the first feature value acquired by the first feature value acquisition unit 11 and the second feature value acquired by the second feature value calculation unit 13 to perform evaluation within each local region. Performs optimization processing to find the optimum solution for a function.

Specifically, the local region optimization processing unit 141 uses the resonance frequency and its gain, which are the first feature values, and the resonance frequency and its gain, which are the second feature values, to perform the evaluation function Find the optimal solution that minimizes Then, in subsequent step SB7, the optimum solution selection unit 142 of the optimization processing unit 140 determines the smallest value among the optimum solutions in each of the local regions as the optimum solution in the search region, and obtains the variables at that time.

At step SB8, the estimation unit 15 estimates the variables at that time as the characteristic values of the controlled object W when the optimization processing unit 140 obtains the optimum solution of the evaluation function in the search area. After that, this flow ends (END).

In the above-described flow, the step SB2 corresponds to the first characteristic value acquisition process, the steps SB6 and SB7 correspond to the optimization process, and the step SB8 corresponds to the estimation process.

By the way, when the optimum solution of the evaluation function is obtained by changing the variables in the entire search area as in the first embodiment, the local solution may be obtained instead of the optimum solution of the evaluation function. FIG. 9 is a diagram showing an example of frequency characteristics obtained using variables when the local solution of the evaluation function is obtained. As shown in FIG. 9, the frequency characteristics obtained using the variables in the case of finding the local solution of the evaluation function are significantly different from the actual measurement results of the frequency characteristics. That is, the variables obtained are different from the actual values.

On the other hand, by dividing the variable search area into a plurality of local areas as in the present embodiment, frequency characteristics close to the measurement result can be obtained as shown in FIG. FIG. 10 is a diagram showing an example of frequency characteristics when the optimum solution of the evaluation function is obtained by the method of the second embodiment. FIG. 10 shows an example in which the spring constants k1 and k3 are not divided, but the viscous damping coefficients c1 to c3 are each divided into five to find the optimum solution.

As described above, the configuration of the present embodiment makes it possible to accurately obtain the optimum solution of the evaluation function in the search area. Therefore, with the configuration of this embodiment, the characteristic values of the controlled object W can be estimated with high accuracy.

In this embodiment, the search area has a plurality of divided local areas. The optimization processing unit 140 changes the second feature value calculated by the second feature value calculation unit 13 by changing the variables of the state equation in each of the plurality of local regions, Obtain the optimal solution of the evaluation function. The estimation unit 15 sets the optimum solution among the optimum solutions of the evaluation function obtained in each local region by the optimization processing unit 140 as the optimum solution of the evaluation function in the search region.

As a result, the optimum solution of the evaluation function can be obtained in each of the local regions obtained by dividing the variable search region into a plurality of regions. Then, by obtaining the optimum solution in the search area from among the optimum solutions in each local area, it is possible to obtain the optimum solution for the entire search area instead of the local optimum solution. Therefore, the optimum solution of the evaluation function in the search area can be obtained with high accuracy. Therefore, the characteristic value of the controlled object W can be obtained with high accuracy.

Moreover, with the above configuration, the characteristic value can be estimated even if the search area is wide, so the characteristic value can be obtained with high accuracy without adjusting the range of the search area and without obtaining a local solution. .

On the other hand, if the variable search area is not divided into a plurality of areas, there is a high possibility that a local solution will be found, so it is necessary to change the search area through trial and error. Also, although it is conceivable to avoid the calculation of the local optimum by increasing the number of iterations of the optimization or by tightening the judgment value of the convergence evaluation of the evaluation function, the calculation takes time. On the other hand, by dividing the search area into a plurality of local areas as in the configuration described above, it is possible to obtain the optimum solution without trial and error or changing the judgment value for convergence evaluation. Therefore, by dividing the search area into a plurality of areas as described above and obtaining the optimum solution of the evaluation function in each local area, the computational load of the characteristic value estimation device 100 can be reduced.

<Embodiment 3>
FIG. 11 is a diagram showing a schematic configuration of a characteristic value estimation device 200 according to Embodiment 3 using functional blocks. The characteristic value estimating device 200 differs in configuration from the characteristic value estimating device 1 of the first embodiment in that it uses data of other frequencies as well as the resonance frequency. In the following description, the same reference numerals are given to the same configurations as in the first embodiment, the description thereof is omitted, and only the configurations different from the first embodiment will be described.

Characteristic value estimation device 200 includes first characteristic value acquisition section 211 , parameter setting section 12 , second characteristic value calculation section 213 , optimization processing section 214 , and estimation section 15 .

The first feature value acquisition unit 211 acquires the resonance frequency and the gain of the controlled object W from the measurement results of the input/output frequency characteristics of the controlled object W, and acquires the gain of the controlled object W at the desired frequency, The gain of the controlled object W at frequencies near the resonance frequency is also obtained. The first feature value acquisition unit 211 has a first acquisition unit 211a and a second acquisition unit 211b.

The first acquisition unit 211a acquires the resonance frequency of the controlled object W and its gain. The second acquisition unit 211b acquires the gain of the controlled object W at the desired frequency, and also acquires the gain of the controlled object W at frequencies near the resonance frequency. In this embodiment, the desired frequency and the neighboring frequencies are predetermined frequencies in the present invention.

In the present embodiment, the resonance frequency and gain acquired by the first acquisition section 211a and the second acquisition section 211b correspond to the first feature value.

The second feature value calculation unit 213 uses the state equation of the controlled object W to obtain the resonance frequency and its gain after changing the variables of the state equation within the search region, and calculates the gain of the controlled object at the desired frequency. The gain of W is obtained, and the gain of the controlled object W at frequencies near the resonance frequency is also obtained. The second feature value calculator 213 has a first calculator 213a and a second calculator 213b.

The first calculation unit 213a uses the state equation of the controlled object W to obtain the resonance frequency and its gain after changing the variables of the state equation within the search area.

Using the equation of state of the controlled object W, the second calculation unit 213b obtains the gain of the controlled object W at the desired frequency, and also obtains the gain of the controlled object W at frequencies near the resonance frequency. In this embodiment, the desired frequency and the neighboring frequencies are predetermined frequencies in the present invention.

In the present embodiment, the resonance frequency and gain calculated by the first calculator 213a and the second calculator 213b correspond to the second characteristic value.

The optimization processing unit 214 changes the variables of the state equation of the controlled object W within the search area, and uses the changed variables to calculate the second feature value calculated by the second feature value calculation unit 213 and the first feature value Using the first feature value acquired by the value acquisition unit 11, optimization processing is performed to find the optimum solution of the evaluation function within the search area. Specifically, the optimization processing unit 214 uses the resonance frequency and gain that are the first feature values and the resonance frequency and gain that are the second feature values to minimize the evaluation function within the search area. Find the optimal solution and find the variables at that time.

The evaluation function is a relational expression between the first feature value and the second feature value. That is, the evaluation function is a relational expression including the square of the difference between the first feature value and the second feature value. Specifically, the evaluation function is the square of the difference between the resonance frequency that is the first feature value and the resonance frequency that is the second feature value, the gain that is the first feature value, and the second feature value. It is a relational expression including the square of the difference with the gain. Specifically, the evaluation function is, for example, the following formula (2).

Here, w1 to w5 are weighting coefficients, fp is the resonance frequency that is the first feature value, fpm is the resonance frequency that is the second feature value, ε is the desired frequency, fpl, fph is a frequency in the vicinity of the resonance frequency, G(fp), G(ε), G(fpl), G(fph) are gains that are first feature values, Gm(fpm), Gm(ε), Gm(fpl) and Gm(fph) are gains that are second feature values. In this embodiment, fpl is a frequency lower than the resonance frequency, and fgh is a frequency higher than the resonance frequency. FIG. 12 shows the relationship of each value in the evaluation function. In the example shown in FIG. 12, the desired frequency is a frequency lower than the resonance frequency.

In equation (2), w3(G(ε)-Gm(ε)) ² performs fitting at the desired frequency, and w4(G(fpl)-Gm(fpl)) ² +w5(G(fph) −Gm(fph)) ² can improve the accuracy of the fitting for resonance damping.

The method of finding the optimum solution using the evaluation function in the optimization processing unit 214 is the same as in the first and second embodiments. Therefore, detailed description of the calculation method for obtaining the optimum solution of the evaluation function is omitted.

With the configuration of the present embodiment, it is possible to improve the accuracy of fitting relating to damping of resonance while performing fitting at the desired frequency, so that the characteristic value of the controlled object W can be easily and accurately estimated. .

The evaluation function may not include w3(G(ε)-Gm(ε)) ² , or w4(G(fpl)-Gm(fpl)) ² +w5(G(fph)-Gm (fph)) ² may not be included. Also, ε may be a frequency higher than the resonance frequency. The evaluation function may be an evaluation function capable of handling a plurality of resonance frequencies like the evaluation function of the first embodiment.

FIG. 13 shows an example of frequency characteristics when the moment of inertia, spring constant, and viscous damping coefficient differ between the measured values and the calculated values. In FIG. 13, the measured values and the calculated values of the resonance frequency and its gain match, while the measured values and the calculated values of the gains of other frequencies are different. In the case shown in FIG. 13, the calculated value can be fitted to the actual measured value by using the above-described equation (2).

In the case of FIG. 13, in addition to W1(G(fp)-Gm(fpm)) ² and W2(fp-fpm) ² in equation (2), w3(G(ε)-Gm(ε )) ² , w4(G(fpl)−Gm(fpl)) ² , and w5(G(fph)−Gm(fph)) ² , an equation including at least one of 2 may be used as the evaluation function. In the case of FIG. 13, W1(G(fp)-Gm(fpm)) ² and W2(fp-fpm) ² are not included in equation (2), and w3(G(ε)-Gm( Expressions including ε)) ² , w4(G(fpl)−Gm(fpl)) ² , w5(G(fph)−Gm(fph)) ² may be used as evaluation functions.

FIG. 14 shows an example of frequency characteristics when the viscous damping coefficient differs between the measured value and the calculated value. In FIG. 14, the measured value and the calculated value of the resonance frequency match, whereas the measured value and the calculated value differ in the gain at the resonance frequency. In the case shown in FIG. 14, the calculated value can be fitted to the measured value by using the following equation (3).

In the case of FIG. 14, in addition to W1(G(fp)-Gm(fpm)) ² and W2(fp-fpm) ² in equation (3), w3(G(fpl)-Gm(fpl )) ² and w4(G(fph)−Gm(fph)) ² may be used as the evaluation function. In the case of FIG. 14, in addition to W1(G(fp)-Gm(fpm)) ² in equation (3), w3(G(fpl)-Gm(fpl)) ² , w4(G( fph)-Gm(fph)) ² may be used as the evaluation function. In the case of FIG. 14, W1(G(fp)-Gm(fpm)) ² and W2(fp-fpm) ² are not included in equation (3), and w3(G(fpl)-Gm( An expression including fpl)) ² and w4(G(fph)-Gm(fph)) ² may be used as the evaluation function.

FIG. 15 shows an example of frequency characteristics when the spring constant differs between the measured value and the calculated value. In FIG. 15, the measured value and the calculated value of the resonance frequency are different from each other, while the measured value and the calculated value of the gain at the resonance frequency match. Also in the case shown in FIG. 15, the calculated value can be fitted to the measured value by using the equation (3).

Also in the case of FIG. 15, in addition to W1(G(fp)-Gm(fpm)) ² and W2(fp-fpm) ² in equation (3), w3(G(fpl)-Gm(fpl) ) ² and w4(G(fph)−Gm(fph)) ² may be used as the evaluation function. In the case of FIG ^. 15, in addition to W1 (G (fp) - Gm (fpm)) ² , w3 (G (fpl) -Gm(fpl)) ² and w4(G(fph)-Gm(fph)) ² may be used as the evaluation function. In the case of FIG. 15, W1(G(fp)-Gm(fpm)) ² and W2(fp-fpm) ² are not included in equation (3), and w3(G(fpl)-Gm( An expression including fpl)) ² and w4(G(fph)-Gm(fph)) ² may be used as the evaluation function.

(Other embodiments)
Although the embodiments of the present invention have been described above, the above-described embodiments are merely examples for carrying out the present invention. Therefore, without being limited to the above-described embodiment, it is possible to modify the above-described embodiment as appropriate without departing from the spirit thereof.

In the second embodiment, the characteristic value estimating device 100 divides each of the viscous damping coefficients c1 to c3 into five without dividing the spring constants k1 and k3 to find the optimum solution of the evaluation function. However, the characteristic value estimation device may divide all of the spring constant and the viscous damping coefficient, or may divide at least one of the spring constant and the viscous damping coefficient.

In each of the above-described embodiments, the controlled object W is a four-inertia system and includes a motor, a torque detector, a test piece, and the like. However, the controlled object W may be a multi-inertia system other than a four-inertia system, such as a three-inertia system or a five-inertia system, or may be a two-inertia system. In addition, the controlled object W may include a configuration other than the above, or may include a shaft system having a configuration other than the above.

In each of the embodiments described above, the first feature value is the resonance frequency and the gain. The second feature value is the resonance frequency and gain. However, the first characteristic value may be only the resonance frequency. The second characteristic value may be only the resonance frequency. Also, the first feature value may include a feature value other than the resonance frequency of the controlled object W and its gain (motor current response, etc.), or may be only the feature value. The second feature value may include a feature value other than the resonance frequency and gain of the controlled object W (motor current response, etc.), or may be only the feature value.

In each embodiment described above, the characteristic value estimation devices 1 and 100 estimate the spring constants k1 and k3 of the object W to be controlled. However, the characteristic value estimation device may estimate other parameters.

In each of the above-described embodiments, the optimization processing units 14 and 140 obtain the minimum value of the evaluation function in the variable search area as the optimum solution. However, the optimization processing section may obtain other values such as the maximum value of the evaluation function in the search area as the optimum solution. In this case, as the evaluation function, instead of the evaluation function of the first embodiment, an evaluation function that can obtain the other value as the optimum solution may be used. The evaluation function used in the optimization processing section may be an evaluation function different from the evaluation functions used in the above embodiments.

In each of the above embodiments, the optimization processing units 14 and 140 set the initial values of the variables to the minimum values of the search area or local area. However, the optimizer may set the initial values of the variables to other values in the search area or local area.

In each of the embodiments described above, an example in which the spring constants k1 and k3 are changed is described as an example of the search area. However, the search area may be an area in which other parameters, which are variables of the state equation of the controlled object, are changed.

INDUSTRIAL APPLICABILITY The present invention can be used for a characteristic value estimation device and a characteristic value estimation method for estimating characteristic values of a controlled object.

1, 100, 200 Characteristic value estimation device 11, 211 First characteristic value acquiring unit 12, 120 Parameter setting unit 13, 213 Second characteristic value calculating unit 14, 140, 214 Optimization processing unit 15 Estimating unit 121 Search region dividing unit 141 Local region optimization processing unit 142 Optimal solution selection unit 211a First acquisition unit 211b Second acquisition unit 213a First calculation unit 213b Second calculation unit W Control target

Claims (5)

A characteristic value estimating device for estimating a characteristic value of a controlled object,
a first feature value acquisition unit that obtains a first feature value of the controlled object at a predetermined frequency from the frequency characteristics of the controlled object;
A second feature value for calculating a second feature value of the controlled object when the variable is changed within the search area at the predetermined frequency using a state equation including a variable that affects the first feature value. a calculation unit;
By changing the variable within the search area, the second feature value calculated by the second feature value calculator is changed, and the relational expression between the first feature value and the second feature value is an optimization processing unit that obtains an optimal solution for a certain evaluation function in the search area;
an estimating unit that estimates the variable when the optimum solution is obtained by the optimization processing unit as the characteristic value;
having
Characteristic value estimator. In the characteristic value estimation device according to claim 1,
The evaluation function is a relational expression including the square of the difference between the first feature value and the second feature value,
The optimization processing unit obtains the value of the evaluation function when the square of the difference between the first feature value and the second feature value is the minimum as the optimal solution,
Characteristic value estimator. In the characteristic value estimation device according to claim 1 or 2,
the first feature value and the second feature value each include at least a gain;
Characteristic value estimator. In the characteristic value estimation device according to any one of claims 1 to 3,
The search area consists of a plurality of local areas,
The optimization processing unit
In each of the plurality of local regions, by changing the variables of the state equation, the second feature value calculated by the second feature value calculator is changed to obtain the optimum solution of the evaluation function. a local area optimization processing unit;
an optimal solution selection unit that selects the optimal solution among the optimal solutions of the evaluation function obtained in each local region by the local region optimization processing unit as the optimal solution of the evaluation function in the search region;
having
Characteristic value estimator. A characteristic value estimation method for estimating a characteristic value of a controlled object,
a first characteristic value acquisition step of obtaining a first characteristic value of the controlled object at a predetermined frequency from the frequency characteristics of the controlled object;
calculating a second characteristic value of the controlled object when the variable is varied within the search area at the predetermined frequency using a state equation including variables that affect the first characteristic value; an optimization processing step of obtaining an optimal solution in the search area of the evaluation function, which is a relational expression between one feature value and the second feature value;
an estimation step of estimating the variable when the optimum solution is obtained in the optimization processing step as the characteristic value;
having
Characteristic value estimation method.

Images