WO2013077007A1 - モータ制御装置 - Google Patents
モータ制御装置 Download PDFInfo
- Publication number
- WO2013077007A1 WO2013077007A1 PCT/JP2012/057217 JP2012057217W WO2013077007A1 WO 2013077007 A1 WO2013077007 A1 WO 2013077007A1 JP 2012057217 W JP2012057217 W JP 2012057217W WO 2013077007 A1 WO2013077007 A1 WO 2013077007A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- model
- target value
- input
- gain
- controller
- Prior art date
Links
Images
Classifications
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
- G05B17/00—Systems involving the use of models or simulators of said systems
- G05B17/02—Systems involving the use of models or simulators of said systems electric
-
- H—ELECTRICITY
- H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
- H02P—CONTROL OR REGULATION OF ELECTRIC MOTORS, ELECTRIC GENERATORS OR DYNAMO-ELECTRIC CONVERTERS; CONTROLLING TRANSFORMERS, REACTORS OR CHOKE COILS
- H02P23/00—Arrangements or methods for the control of AC motors characterised by a control method other than vector control
- H02P23/0004—Control strategies in general, e.g. linear type, e.g. P, PI, PID, using robust control
-
- H—ELECTRICITY
- H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
- H02P—CONTROL OR REGULATION OF ELECTRIC MOTORS, ELECTRIC GENERATORS OR DYNAMO-ELECTRIC CONVERTERS; CONTROLLING TRANSFORMERS, REACTORS OR CHOKE COILS
- H02P23/00—Arrangements or methods for the control of AC motors characterised by a control method other than vector control
- H02P23/12—Observer control, e.g. using Luenberger observers or Kalman filters
Definitions
- the present invention relates to a motor control device for driving an industrial machine.
- the motor control device is desired to follow the motor position and speed with high speed and high accuracy with respect to the target value such as the position target value and the speed target value, and in order to realize this, feed forward control and feedback control are performed. Combined two-degree-of-freedom control is used.
- a control technique that prefetches and uses not only the current target value but also the future target value may be used.
- it is also necessary to suppress residual vibration at the time of stoppage caused by mechanical resonance, etc., so that the signal component at the vibration frequency (resonance frequency) of industrial machinery is reduced.
- vibration suppression is realized by vibration suppression control that designs a feedforward control system.
- Patent Document 1 there is an electric motor control device that performs predictive control using a target value up to two steps or four steps ahead for predictive control that required a target value up to M (M> 4) steps ahead. It is disclosed.
- the control calculation is simplified by using only the target values up to two steps or four steps ahead, and the speed feed forward and torque feed forward are calculated from these values and used for control. As a result, the overshoot during acceleration / deceleration is reduced and the followability at a constant speed is improved.
- the servo control device described in Patent Document 2 has a changeover switch, and when the changeover switch is ON, feedback control using a predictive controller as a feedback controller or two-degree-of-freedom control is used. When the switch is OFF, the performance is improved by switching the feedback control of only the position proportional control.
- the changeover switch is turned on when the controlled object requires high-accuracy followability, and the changeover switch is turned off when the target command increment value is changing. Further, in order to make the control input before and after switching continuous, the control input before and after switching is linearly interpolated by a filter function unit.
- the present invention has been made in view of the above-described problems, and uses not only the current target value but also the past target value, and the amount of change in the control input and control input for any position command.
- An object of the present invention is to provide a motor control device that allows a motor to follow a target value with high speed and high accuracy and that does not excite vibrations when stopped.
- the present invention provides a model output representing a desired operation of the control object based on a target value to be followed by the control output of the control object including the motor,
- a normative model unit that generates a model input that drives the controlled object in a desired operation
- a feedback control unit that inputs the control output and the model output and generates a feedback input that causes the control output to follow the model output
- a model input adder that adds the model input and the feedback input to generate a control input to the control target
- the reference model unit includes a current value of the target value
- a target value storage unit that holds a value and one or more past values of the target value as a target value vector, and simulates the characteristics of the control target
- a mathematical model that generates the model output and state variable based on a Dell input
- a model controller that generates the model input based on the target value vector and the state variable
- a target value vector and the state variable that determines the
- the present invention in response to variously changing commands, not only the current target value but also the past target value is used, and the characteristics of the controller are changed while the motor is being driven. It is possible to obtain a motor control device that causes the motor to follow a target value and that does not excite vibrations when stopped.
- FIG. 1 is a block diagram showing a configuration of a motor control apparatus according to Embodiment 1 of the present invention.
- FIG. 2 is a diagram showing a two-inertia system model which is an example of a mathematical model of the motor control device according to Embodiment 1 of the present invention.
- FIG. 3 is a block diagram showing the configuration of the model controller of the motor control apparatus according to Embodiment 1 of the present invention.
- FIG. 4 is a block diagram showing the configuration of the motor control device according to Embodiment 2 of the present invention.
- FIG. 5 is a diagram showing an example of the maximum output allowable set of the motor control device according to the second embodiment of the present invention.
- FIG. 6 is a diagram illustrating an operation example of the motor control device according to the second embodiment of the present invention.
- FIG. 7 is a block diagram showing the configuration of the motor control apparatus according to Embodiment 3 of the present invention.
- FIG. 8 is a block diagram showing an example of model controller candidates of the motor control apparatus according to Embodiment 3 of the present invention.
- FIG. 9 is a block diagram showing an example of model controller candidates of the motor control apparatus according to Embodiment 3 of the present invention.
- FIG. 10 is a block diagram showing an example of model controller candidates of the motor control device according to Embodiment 3 of the present invention.
- FIG. 1 is a block diagram showing a configuration of a motor control device 100 according to Embodiment 1 of the present invention.
- the same reference numerals indicate the same or corresponding parts.
- the motor control device 100 includes a reference model unit 1, a feedback control unit 2, and a model input adder 3.
- the target value r (i) for the control output y (i) of the controlled object 4 output by the detector 5 representing the position and speed of the machine to be driven or the position and speed of the motor to drive the machine is externally supplied to the motor control device. 100, and the motor control device 100 controls the control input u (i) representing the torque, current, etc. of the motor so that the control output y (i) from the detector 5 of the controlled object 4 follows the target value r (i). ).
- the calculated control input u (i) is output from the motor control device 100 to the control object 4. Note that i used for each symbol represents the number of steps.
- the control object 4 includes a mechanical load and a motor such as a rotary motor or a linear motor that drives the mechanical load.
- a control input u (i) which will be described later, to the motor to be controlled 4, the machine load is caused to perform a desired operation.
- the detector 5 is an encoder, a linear scale, etc., and detects the current position information and speed information of the rotary motor and linear motor constituting the control object 4 or the current position information and speed information of the mechanical load, and Is output to the feedback control unit 2 as a control output y (i).
- control target 4 and the detector 5 are collectively regarded as a control target including a motor, the input to the control target is the control input u (i), and the output from the control target is the control output y (i). ing.
- the reference model unit 1 to which the target value r (i) is input outputs a model output yM (i) representing an ideal operation waveform of the control target 4 to the feedback control unit 2, and an ideal operation for the control target 4.
- the model input uM (i) for causing the model input to be output to the model input adder 3.
- the normative model unit 1 to which the target value r (i) is input stores and holds the target value r (i) for a predetermined time, and currently inputs one or a plurality of past target values stored and held.
- a target value storage unit 15 that outputs the target value r (i) as a target value vector rvec (i) is provided.
- the normative model unit 1 stores and holds the model input uM (i) for a predetermined time, and outputs the stored value as the past model input uM ′ (i), and the model input uM (i ) As an input and a model output 14 that simulates the characteristics of the controlled object 4 and outputs a model output yM (i) and a state variable xM (i). Further, the normative model unit 1 includes a target value vector rvec (i) output from the target value storage unit 15, a state variable xM (i) output from the mathematical model 14, and a past model input output from the model input memory 13. uM ′ (i) is used as an input, and a model controller determination unit 11 that determines a model controller to be actually used from a plurality of predetermined model controller candidates based on these input values is provided.
- the normative model unit 1 receives the target value vector rvec (i), the state variable xM (i), and the past model input uM ′ (i), and uses the model controller determined by the model controller determination unit 11. And a model controller 12 that calculates a model input uM (i) that causes the model output yM (i) to follow the target value r (i) and outputs the model input 14 to the model input adder 3.
- the state variable xM (i) represents the internal state of the mathematical model 14, and is an n-order (n is 1 or more) numeric vector.
- the model output yM (i) is a part of the state variable xM (i) of the mathematical model 14 corresponding to the physical quantity that can be actually measured from the controlled object 4, such as the position and speed of the motor, or both.
- the dimension is described as m.
- the feedback control unit 2 receives the model output yM (i) output from the reference model unit 1 and the control output y (i) output from the detector 5, and the control output y (i) is the model output yM (i). ), The feedback control input uFB (i) is calculated, and the calculated feedback control input uFB (i) is output to the model input adder 3. That is, the feedback control unit 2 receives the model output yM (i) and the control output y (i), calculates the difference between these values, and outputs the calculation result as the output deviation e (i).
- a feedback controller 22 that outputs a feedback input uFB (i) so as to follow.
- the model input adder 3 adds the feedback input uFB (i) output from the feedback control unit 2 and the model input uM (i) output from the reference model unit 1 and uses the sum as the control input u (i). And output to the control object 4. Then, the motor attached to the control object 4 is driven by the control input u (i), and the control output y (i) of the control object 4 and the target value r (i) are matched. As a result, the control object 4 follows the target value and performs a desired operation.
- a target value r (i) is input to the target value storage unit 15, the target value r (i) is stored and held for a predetermined time, and one or more past target values stored and held are currently input.
- the target value r (i) is output to the model controller determination unit 11 and the model controller 12 as a target value vector rvec (i).
- the target value storage unit 15 includes one to P target value memories 151 (P> M), and the target value r (input to the target value storage unit 15 is input to these target value memories 151). i) is input, and the target value memory 151 stores and holds the target value r (i) for 1 to M steps. In this case, all the target values input for M ⁇ Ts time are stored and held.
- Ts is one sampling time. That is, assuming that the current target value is r (i), the target value memory 151 sets the values of r (i), r (i-1)... R (i ⁇ M ⁇ 1) for one step. Keep in memory.
- r (i-1) represents a target value one step before (time one sampling time Ts before the current time)
- r (i-2) represents a target value two steps before.
- One step number corresponds to the sampling time Ts
- storing the target value r (i) for M steps corresponds to storing and holding the target value r (i) for M ⁇ Ts time. .
- the past target value r (i ⁇ 1)... R (i ⁇ M) output from the target value memory 151 and the current target value r (i) are simultaneously collected to obtain a target value vector rvec (i). Output as shown in equation (1).
- the model input uM (i ⁇ 1) of the previous step is used as the past model input uM ′ (i), and the target value memory 151 stores and holds the target value for two steps.
- the mathematical model 14 calculates the state variable xM (i + 1) and the model output yM (i) of the next step based on the input model input uM (i) by the discrete time state equation of the following equation (2). .
- I a matrix representing the characteristics of the mathematical model 14.
- A, B, C, and D shown in the above equation (4) are determined so as to represent the characteristics of the controlled object 4.
- the mathematical model 14 does not have to represent all the characteristics of the controlled object 4, and only simulates the characteristics necessary for the control output y (i) to follow the target value r (i) at high speed and high accuracy. .
- FIG. 2 is a diagram showing a two-inertia system model that is an example of the mathematical model 14 of the motor control device 100 according to Embodiment 1 of the present invention.
- a two-inertia system model in which the motor as shown in FIG. Consider the case of simulating low resonance / anti-resonance characteristics.
- a vibrational model having resonance / antiresonance such as a two-inertia model is used for the mathematical model 14, so that the motor control has a high vibration suppression effect and is high speed and high accuracy.
- An apparatus can be obtained.
- the position and speed of the motor attached to the controlled object 4 can be measured by using the detector 5.
- pM is the motor position (rotation angle)
- ⁇ M is the motor rotation speed
- pL is the mechanical load position
- ⁇ L is the mechanical load speed
- JM is the motor inertia moment
- JL is the mechanical load inertia moment
- TM is the motor torque
- km is the spring constant
- cm is the viscosity coefficient
- yM is the model output of the continuous system
- d / dt is the derivative with respect to time.
- the model resonance frequency ⁇ p and the model anti-resonance frequency ⁇ z calculated by the following equations (7) and (8) coincide with the resonance frequency / anti-resonance frequency of the lowest frequency of the controlled object 4. You may decide so. By determining in this way, the two-inertia system model can simulate the characteristics of the controlled object 4.
- the mathematical model 14 is a discrete-time state equation as shown in the equation (2), the equation (5) and the equation (6) are discretized to obtain a discrete-time state equation, and the mathematical model 14 is obtained. .
- e AcTs and e Ac (Ts ⁇ ) represent exponential functions of AcTs and Ac (Ts ⁇ ), respectively, and Ts represents a sampling time.
- the numerical formula model 14 is the formula (9), the formula (10), the formula (11), the formula (12), the state variable xM (i), the model input.
- uM (i) and model output yM (i) are not limited to Equation (13).
- the model controller 12 includes a model sub-controller 121 and a model input change amount adder 122. Based on the target value vector rvec (i), the model input uM (i ⁇ 1) one step before, and the state variable xM (i) from the mathematical model 14, the model controller determination unit 11 sets a plurality of predetermined values. A model controller is determined from the candidates. The model controller determined by the model controller determination unit 11 is input to the model sub-controller 121. The model sub-controller 121 is present in the state variable xM (i) at the target value r (i) existing in the target value vector rvec (i) based on the model controller determined by the model controller determination unit 11.
- the model input change amount ⁇ uM (i) is calculated so that the model output yM (i) follows. Then, the calculated model input change amount ⁇ uM (i) is output to the model input change amount adder 122.
- the model input change amount adder 122 adds the model input change amount ⁇ uM (i) output from the model sub-controller 121 and the model input uM (i ⁇ 1) of the previous step and adds the sum to the model input uM.
- the data is output to the model input memory 13, the mathematical model 14, and the model input adder 3.
- the model sub-controller 121 calculates the model input change amount ⁇ uM (i) by adding the state feedback and the offset.
- the model sub-controller 121 reads the target value vector rvec (i), the model input uM (1 step before)
- the model input change amount ⁇ uM (i) is calculated based on i-1), multiplication of the state variable xM (i) of the mathematical model 14 and the model gain Ki, and addition of the model offset Gi.
- the calculation of the model sub-controller 121 can be expressed by the following equations (14), (15), and (16).
- vr (i) is a target value speed and ar (i) is a target value acceleration, which are values corresponding to the first-order difference and second-order difference of the target value r (i), respectively.
- Kri is a target value model gain
- Kvri is a target value speed model gain
- Kari is a target value acceleration model gain.
- the model offset Gi may not be required, but is introduced for use later.
- a plurality of candidates for the model gain Ki, the model offset Gi, the target value speed model gain Kvri, and the target value acceleration model gain Kari are determined in advance by a method to be described later, and the model controller determining unit 11 determines a plurality of candidates. And the determined value is used in equation (14).
- the term relating to the target value r (i) is shown separately, but in reality, the model input change amount is calculated from the deviation between the target value r (i) and the state variable xM (i) to be followed. ⁇ uM (i) may be generated.
- the target value model gain Kri is automatically determined from the model gain Ki. Is.
- the model input change amount adder 122 adds the model input change amount ⁇ uM (i) and the one-step previous model input uM (i ⁇ 1) from the model sub-controller 121 as represented by the following equation (17). Then, the sum is output as a model input uM (i) to the model input memory 13, the mathematical model 14, and the model input adder 3. That is, the model input uM (i) is calculated by integrating the model input change amount ⁇ uM (i).
- the model controller 12 calculates the model input change amount ⁇ uM (i) by the model sub-controller 121 by performing the calculation as described above, and the model input change amount ⁇ uM by the operation of the model input change amount adder 122.
- the model input uM (i) is output by integrating (i).
- the model controller determination unit 11 and the model controller 12 can know the value of the model input change amount ⁇ uM (i) in addition to the model input uM (i).
- the model sub-controller 121 can be determined from a plurality of candidates so that uM (i) and the model input change amount ⁇ uM (i) do not exceed a predetermined value.
- the model controller determining unit 11 causes the model output yM (i) output from the mathematical model 14 to follow the target value r (i) with high speed and high accuracy, or the model input change amount ⁇ uM (i) is predetermined.
- the model input uM (i) can be calculated without causing the model input uM (i) to be discontinuous even if the model sub-control 121 is switched to another candidate.
- model input change amount ⁇ uM (i) is regarded as a new input, and a discrete time state equation in which the model input change amount adder 122 and the model input memory 13 are combined with the mathematical model 14 is expressed by the following equation (18). be able to.
- I represents a unit matrix
- 0 represents a zero matrix
- the number of rows and columns is appropriately determined.
- a discrete time state equation that also takes into account the storage of the target value r (i) by the target value storage unit 15 can be expressed by the following equation (19). .
- Equation (19) is used by the model controller determining unit 11 to calculate candidates for a plurality of model controllers 12 or model sub-controllers 121 to be determined in advance, which will be described later.
- the offset Gi, the target value model gain Kri, the target value speed model gain Kvri, and the target value acceleration model gain Kari can be designed.
- the specific calculation method of the model controller 12 is not limited to the one shown in the configuration of FIG. 3, and a calculation equivalent to the above is possible with various configurations. That is, the model controller 12 is based on the target value vector rvec (i), the state variable xM (i) of the mathematical model 14 and the model input uM (i ⁇ 1) one step before which is the output of the model input memory 13. By performing the above calculation, a calculation equivalent to the above can be performed, and a calculation in consideration of the model input change amount ⁇ uM (i) can be performed.
- the model controller determination unit 11 determines the model gain Ki, the model offset Gi, the target value.
- the model gain Kri, the target value speed model gain Kvri, and the target value acceleration model gain Kari are determined from a plurality of predetermined candidates.
- the target value model gain Kri is automatically determined from the model gain Ki as described above.
- the model input uM (i ⁇ 1) one step before, the model gain Ki, the target value speed model gain Kvri, the target value acceleration A gain map describing the correspondence between the model gain Kari, the target value vector rvec (i), the state variable xM (i), the model input uM (i ⁇ 1) one step before, the model A method of determining the model gain Ki, the model offset Gi, the target value speed model gain Kvri, and the target value acceleration model gain Kari using an offset map that describes the correspondence relationship with the offset Gi will be described.
- a model gain Ka that does not excite the vibration of the control output y (i) to the target value r (i) and the control output y (i) quickly follows the target value r (i).
- the model offset Ga, the target value speed model gain Kvra, the target value acceleration model gain Kara, and the control output y (i) are not excited with the target value r (i), but the control output y (i) is not the target value r (
- a model gain Kb, a model offset Gb, a target value speed model gain Kvrb, and a target value acceleration model gain Karb for obtaining a response following i) slowly are prepared. It is assumed that ⁇ Ka, Kvra, Kara, Kb, Kvrb, Karb ⁇ are saved in the gain map, and ⁇ Ga, Gb ⁇ is saved in the offset map.
- a large control input u (i) or model uM is used to quickly follow the control output y (i) to the target value r (i) without exciting vibration.
- the model input change amount ⁇ uM (i) is required, and if the tracking of the control output y (i) with respect to the target value r (i) is delayed, a small control input u (i) or model input uM (i), The model input change amount ⁇ uM (i) is sufficient. Therefore, the model controller determination unit 11 normally determines ⁇ Ka, Kvra, Kara ⁇ from the gain map and Ga from the offset map so that the control output u (i) quickly follows the target value r (i). Output to the model controller 12.
- the model controller determination unit 11 determines ⁇ Kb, Kvrb, Karb ⁇ determines Gb from the offset map and outputs it to the model controller 12.
- the control output y (i) oscillates to the target value r (i). If the control input u (i), the model input uM (i), and the model input change amount ⁇ uM (i) are likely to exceed predetermined values, the target value r () of the control output y (i) Although the follow-up to i) is delayed, there is no vibration and a response is obtained in which the model input uM (i) and the model input change amount ⁇ uM (i) do not exceed predetermined values.
- Various gains and offset candidates ⁇ Ka, Ga, Kvra, Kara ⁇ , ⁇ Kb, Gb, Kvrb, Karb ⁇ are determined by the pole placement method, loop shaping, etc. for the discrete time state equation of Equation (19).
- the design is possible by using the control system design method, and the present embodiment is not particularly limited.
- the number of gain sets is limited to two sets of gains in which the control output y (i) quickly follows the target value r (i) and the gain in which the control output y (i) follows the target value r (i) slowly. There may be three or more sets.
- the model predictive control as follows, the values and determination conditions of a plurality of model gains Ki, model offset Gi, target value speed model gain Kvri, target value acceleration model gain Kari are systematically determined. Can be determined.
- a piecewise affine state feedback control system is designed using model predictive control as a control method not limited to a motor control device, thereby minimizing a certain evaluation function while satisfying constraints.
- a scheme has been proposed.
- the piecewise affine state feedback control system calculates the input applied to the machine based on the multiplication of the state of the machine to be controlled, the gain, and the addition of the offset, and these gain and offset are It is a control system that switches according to the state of.
- a piecewise affine state feedback control system is designed in a portion corresponding to the feedback controller 22 in the present embodiment, it is necessary to measure or estimate the state of the machine to be controlled. There is. Therefore, a large number of sensors are required or an observer needs to be constructed, which is not practical.
- the model input change amount ⁇ uM (i) is considered, and the past target values r (i ⁇ 1) and r (i ⁇ 2) output from the target value storage unit 15 are also considered. Therefore, the discrete time state equation represented by the equation (19) is used as a prediction model used when designing the controller.
- control constraint is that the model input uM (i) and the model input change amount ⁇ uM (i) do not exceed a predetermined value in each step. That is, it is assumed that a control constraint expressed by the following equation (20) is imposed.
- uMmax represents the maximum absolute value of the model input uM (i)
- ⁇ uMmax represents the maximum absolute value of the model input change amount ⁇ uM (i).
- the evaluation function is represented by the following expression (21).
- N is called a horizon, is a parameter that determines how much the future is predicted, and is one design parameter.
- Q, R, and PN are weights and are design parameters.
- the expression (22) does not show the terminal variable constraint set of state variables and the model input constraint set, but the effects of the present embodiment can be obtained without using them. Also, the above reference material shows that stability can be guaranteed by the input obtained by solving the problem of Equation (22) by appropriately setting the terminal constraint set of state variables and the terminal constraint set of model inputs. Yes. Therefore, by using the termination constraint set of the state variable and the termination constraint set of the model input, it is possible to prevent the model input from diverging and the operation from becoming unstable even in the motor control device of the present embodiment.
- xM (i) is a state variable at the i-th step
- uM (i ⁇ 1) is a model input one step before
- v is an input sequence [ ⁇ uM (i), ⁇ uM (i + 1),. , ⁇ uM (i + N ⁇ 1)] and auxiliary variables.
- xM (i) is determined from a plurality of candidates according to the model input uM (i-1) one step before, the target value speed vr (i), and the target value acceleration ar (i).
- a determination condition Pi for determining the model gain Ki, the model offset Gi, the target value speed model gain Kvri, and the target value acceleration model gain Kari is obtained from a plurality of candidates.
- model input uM (i) can be calculated.
- the condition Pi is automatically designed.
- the determination condition Pi includes a plurality of candidates for the model gain Ki, target value speed model gain Kvri, target value acceleration model gain Kari and a gain map indicating the determination condition, and an offset map indicating the plurality of candidates for the model offset Gi and the determination condition. It can be divided into
- the model controller determination unit 11 has a gain map and an offset map obtained as a result of solving the minimization problem of the equation (22) or the equation (23) in advance, and the state variable xM of the equation model 14 at each sampling time.
- the previous model input uM (i-1) and the target value vector vrec (i) are input, the state variable xM (i), the previous model input uM (i-1), and the target value
- the model gain Ki, the model offset Gi, the target value A speed model gain Kvri and a target value acceleration model gain Kari are determined. Further, the determined model gain Ki, model offset Gi, target value speed model gain Kvri, and target value acceleration model gain Kari are output to the model controller 12.
- the model controller 12 determines the state variable xM (i) of the mathematical model 14 and the model input uM (i ⁇ 1) one step before and the model gain according to the equations (24) and (26).
- the model input uM (i) is calculated based on the sum of the value obtained by adding all the products of the target value acceleration model gain Kari and the model offset Gi.
- the model input uM (i) and the model input change amount ⁇ uM (i) in Expression (24), Expression (25), and Expression (26) are derived as a result of the minimization problem in Expression (22) or Expression (23). Therefore, the model input uM (i) and the model input change amount ⁇ uM (i) satisfying the control constraint (20) of the minimization problem are derived. That is, when the absolute values of the model input uM (i) and the model input change amount ⁇ uM (i) are equal to or greater than a predetermined value, the model gain Ki, the model offset Gi, the target value speed model gain Kvri, and the target value acceleration model gain Kari switches.
- a gain map and an offset map that describe candidate values of the model gain Ki, model offset Gi, target value speed model gain Kvri, target value acceleration model gain Kari and conditions for switching these values are automatically calculated. That the model input uM (i) does not exceed a predetermined value means that the control constraint of the control input is satisfied, and that the model input change amount ⁇ uM (i) does not exceed a predetermined value. Even if the model gain Ki, the model offset Gi, the target value speed model gain Kvri, and the target value acceleration model gain Kari are changed based on the gain map and the offset map generated from the determination condition Pi, the model input uM (i) does not change rapidly. It means that.
- model offset Gi is derived as a result of solving the minimizing problems (22) and (23), only the model gain Ki, the target value speed model gain Kvri, and the target value acceleration model gain Kari are used.
- a control system that obtains a high-speed and high-accuracy response in the sense that the evaluation function of Expression (21) is made smaller can be realized by adding the model offset Gi than in the case of doing so.
- model input change amount ⁇ uM (i) is determined by determining model controller 12 from a plurality of predetermined candidates using past model input uM ′ (i).
- the model controller 12 can be determined in consideration of changes in the control input. That is, the model input change amount ⁇ uM (i) can be used as a new input in order to calculate the model input uM (i) using the past model input uM ′ (i).
- the model controller determining unit 11 appropriately determines the model controller 12, not only the model input uM (i) but also the model input change amount ⁇ uM (i) does not exceed a predetermined value,
- the control output y (i) follows the target value r (i) quickly and without vibration. That is, it is possible to obtain a high-speed and high-precision motor control device while preventing torque saturation (current saturation), torque change rate saturation, and voltage saturation closely related to torque change rate saturation. Even if the model controller 12 is switched to improve the control performance, it is possible to prevent the model input uM (i) from changing suddenly.
- the model controller determines the model controller so that the model input becomes smaller than a predetermined value, the occurrence of torque saturation can be prevented and the model controller can be determined so that the amount of change in the model input becomes smaller than the predetermined value.
- the past model input uM ′ (i) is fed back to the model controller 12 to be equivalent to the discrete time state equations of Equations (18) and (19) using the model input variation ⁇ uM (i) as a new input. Can be converted. This makes it possible to consider the model input variation ⁇ uM (i) in the minimization problem (22).
- the target value storage unit 15 By the target value storage unit 15, the target value speed vr (i) and the target value acceleration ar (i) Can be taken into account when calculating the model input change amount ⁇ uM (i), and a control system in which the control output y (i) follows the target value r (i) earlier than when these values are not used can be designed. .
- the model controller is generated and controlled to switch the model controller using only the current target value. It becomes possible to obtain a motor control device with higher response.
- the model controller determination unit 11 determines the model gain Ki, the model offset Gi, the target value according to the past model input uM ′ (i), the state variable xM (i) of the mathematical model 14 and the target value vector rvec (i).
- the model input uM (i) and the model input change amount ⁇ uM (i) do not exceed predetermined values and a high-speed and high-accuracy response can be obtained.
- a control device can be realized. That is, it is possible to obtain a motor controller with high response by using the offset amount in addition to the model gain.
- a gain map and an offset map that describe candidates and determination conditions for the model gain Ki, the model offset Gi, the target value speed model gain Kvri, and the target value acceleration model gain Kari are: It is requested offline in advance.
- the on-line calculation uses the target value vector rvec (i) by referring to the gain map and offset map to calculate the formula (25) for determining the model gain Ki, model offset Gi, target value speed model gain Kvri, Only the calculation of the model input uM (i) by (24) and Expression (26) is performed.
- online model prediction control that calculates the model input change amount ⁇ uM (i) by calculating the minimization problem of Expression (22) at each step. It is possible to shorten the calculation time. Therefore, it is possible to considerably reduce the amount of calculation on-line, the sampling time can be made faster, and the performance of the control system can be improved.
- the multiparametric linear programming problem of Equation (23) is merely replaced with a multiparametric quadratic programming problem.
- the model input uM (i ⁇ 1) one step before, and the target value vector rvec (i) the model gain Ki and the model offset Gi
- a map Pi describing candidate values and determination conditions for each gain for switching between the target value speed model gain Kvri and the target value acceleration model gain Kari is output.
- the model controller determining unit 11 uses the gain map to switch the target value model gain Kri, the target value speed model gain Kvri, and the target value acceleration model gain Kari, or the gain map and the offset map. Is used to determine the target value model gain Kri, the model offset Gi, the target value speed model gain Kvri, and the target value acceleration model gain Kari. However, the state variable xM (i) of the mathematical model 14 and the target value vector The model gain Ki, the model offset Gi, the target value speed model gain Kvri, and the target value acceleration model gain Kari may be determined using a function having rvec (i) and the past model input uM ′ (i) as variables.
- control constraint is not limited thereto.
- control constraints relating to the state variable xM (i), model input uM (i), and model input change amount ⁇ uM (i) of the mathematical model 14 may be convex constraints.
- a control constraint may be imposed on the state variable xM (i) of the mathematical model 14.
- the model gain Ki When determining the model gain Ki, the target value speed model gain Kvri, the target value acceleration model gain Kari, or the model gain Ki, the model offset Gi, the target value speed model gain Kvri, and the target value acceleration model gain Kari It is not necessary to use all the state variables xM (i), target value vector rvec (i), and past model input uM ′ (i) of the model 14, and only a part of these pieces of information may be used.
- the model gain or the model gain and the model offset so that the model input becomes smaller than the predetermined value, the occurrence of torque saturation can be prevented, and the change amount of the model input is less than the predetermined value.
- the model gain or the model gain and the model offset so as to decrease, the occurrence of voltage saturation can be prevented, and the shock when the model controller is switched can be reduced.
- the model input uM (i ⁇ 1) one step before is used as the past model input uM ′ (i).
- the model input uM (i ⁇ 1) is not necessarily one step before. It is easy to give the same effect using the model input, and the same effect as in the present embodiment can be obtained even if a past model input other than the model input one step before is used as the past model input.
- the model input uM (i) is generated using the target value speed vr (i) and the target value acceleration ar (i).
- the model input uM (i) is not used without using these values. It may be generated. That is, the model gain Ki and the target value model gain Kri automatically determined thereby are determined according to the target value vector Kvri, the state variable xM (i), and the model input uM (i-1) one step before.
- the target value speed model gain Kvri and the target value acceleration model gain Kari may be always set to 0, and the equations (14) and (24) may be calculated. Even in this method, the same effect as in the first embodiment can be obtained.
- the model controller 12 includes the model sub-controller 121, calculates the model input change amount ⁇ uM (i), and inputs the model input change amount ⁇ uM (i) one step before the model input uM ( If the model input uM (i) was calculated by adding i-1), but it is not necessary to consider the model input change amount control constraint
- ⁇ ⁇ uMmax, the model input change
- the model input uM (i) may be calculated directly without calculating the quantity ⁇ uM (i). That is, the model input uM (i) may be calculated using the following equation (27).
- model gain Ki, model offset Gi, target value model gain Kri, target value of equation (27) is changed by changing the minimization problem of equation (22) to the minimization problem of equation (28) shown below.
- a gain map and an offset map describing conditions for determining the speed model gain Kvri and the target value acceleration model gain Kari are obtained. That is, by generating a gain map and an offset map so as to minimize the following evaluation function, an appropriate gain map and offset map can be automatically obtained.
- the motor control device 100 by configuring the motor control device 100 as described above, it is possible to reduce the calculation amount to the control target 4 such as the motor torque with a small amount of calculation with respect to various target values. Controller characteristics can be automatically determined without making the control input discontinuous. As a result, it is possible to make the control output follow the target value at high speed and with high accuracy while restricting the absolute value or amount of change of the control input to the control target 4.
- the motor control apparatus 200 detects the target value r (i) for the control output y (i) indicating the position and speed of the machine to be driven, and the detector 5.
- the control output y (i) of the control object 4 representing the position and speed of the motor is input.
- the motor control device 200 outputs the control input u (i) such as the motor torque and current to the controlled object 4 so that the control output y (i) from the detector 5 follows the target value r (i). To do.
- the motor control device 200 includes a reference model unit 1A, a feedback control unit 2, and a model input adder 3.
- the motor control device 200 is connected to a control target 4 including a mechanical load and a motor such as a rotary motor that drives the mechanical load, and a detector 5 such as an encoder. Since the feedback control unit 2, the model input adder 3, the control target 4, and the detector 5 are the same as those in the first embodiment, the description thereof is omitted.
- the target value r (i) is input to the reference model unit 1A, and a model output yM (i) indicating an ideal operation waveform of the control target 4 is output to the feedback control unit 2. Further, the reference model unit 1A uses the model input adder as the model input uM (i) for driving the control target 4 so that the control output y (i) output from the detector 5 follows the target value r (i). Output to 3.
- the normative model unit 1A stores and holds the target value r (i) for a predetermined time, and the current target value r (i) to which one or more past target values stored and held are input.
- a target value storage unit 15 that outputs a target value vector rvec (i) together, a model controller determination unit 11A that determines a model controller from a plurality of predetermined candidates, and a model output yM (i) as a target value It has a model controller 12 that follows r (i), a model input memory 13 that stores the model input uM (i), and a mathematical model 14 that simulates the characteristics of the controlled object 4.
- the model input memory 13, the mathematical model 14, and the target value storage unit 15 are the same as those in the first embodiment, the description thereof is omitted.
- the operation of the model controller determination unit 11A is different from the model controller determination unit 11 of the motor control device 100 according to the first embodiment. That is, in the present embodiment, the model controller determining unit 11A determines a plurality of candidates designed in advance by pole arrangement or the like.
- the model controller determination unit 11A inputs the target value vector rvec (i), the one-step previous model input uM (i ⁇ 1), and the state variable xM (i) of the mathematical model 14, and the model input uM (I) A plurality of model gains Ki, target value model gain Kri, target value speed model gain Kvri, target value designed in advance so that the value of the model input change amount ⁇ uM (i) does not exceed a predetermined value.
- a set is determined from the acceleration model gain Kari, and those gains are output to the model controller 12. In the present embodiment, it is assumed that the model controller determination unit 11A always determines the model offset Gi to be 0.
- model gain Ki By predetermining a plurality of model gain Ki, target value model gain Kri, target value speed model gain Kvri, and target value acceleration model gain Kari, it is possible to reduce the amount of calculation online, and to reduce the sampling time. It is possible to make it as soon as possible. Also in the present embodiment, when the model input change amount ⁇ uM (i) is generated from the deviation between the target value r (i) and the state variable xM (i) to be followed, the target value model gain Kri is the model gain. It is automatically determined from Ki.
- the maximum output allowable set O ⁇ i is used to determine the model gain Ki, the target value speed model gain Kvri, and the target value acceleration model gain Kari.
- the maximum output allowable set O ⁇ i is a set that collects the conditions under which the control system (closed loop system) does not break the control constraints.
- the calculation method is “Hirata, Fujita,“ Restrictions on linear discrete-time systems with external inputs. "Analysis of Conditions", JSME C, 118-3, 384 / 390-, 1998 ".
- the constraint that the model input uM (i) and the model input change amount ⁇ uM (i) described by Expression (20) do not exceed a predetermined value is considered as the control constraint.
- the maximum output allowable set O ⁇ i will be described.
- the motor control device 200 includes the target value storage unit 15, and the model controller 12 included in the normative model unit 1 ⁇ / b> A is one step before as shown in FIG. 3.
- the model input uM (i-1) is fed back to the model sub-controller 121, and the model input change amount adder 122 adds the model input change amount ⁇ uM (i) and the model input uM (i-1) of the previous step.
- the model input uM (i) is generated. Therefore, the discrete time state equation represented by the equation (19) is considered by combining the target value storage unit 15, the model input memory 13, the mathematical expression model 14, and the model input change amount adder 122.
- the model sub-controller 121 uses equation (27) to state the state variable xM (i) of the mathematical model 14, the model input uM (i ⁇ 1) one step before, and the target value r (i ), Calculating the model input change amount ⁇ uM (i) from the target value speed vr (i) and the target value acceleration ar (i) as shown in the following equation (29).
- the model controller determination unit 11A since the model controller determination unit 11A always determines the model offset Gi as 0, the term of the model offset Gi is excluded from the equation (29).
- Equation (30) represents the closed loop system of the reference model portion 1A. If the state variable xM (i) of the mathematical model 14 at the i step is determined, the target value predicted values ⁇ r (i + 1), ⁇ r (i + 2) ... after the i step and the target value r (i + 1), r (i + 2)..., target value speed prediction value ⁇ vr (i + 1), ⁇ vr (i + 2) ... and target value speed vr (i + 1), vr (i + 2) ..., target value acceleration prediction value ⁇ Assuming that ar (i + 1), ⁇ ar (i + 2)... and target value accelerations ar (i + 1), ar (i + 2)... are equal, that is, the following equation (31) holds:
- the state variable xM (i + l) (l> 0), model input uM (i + l) (l> 0), target value r (i + l) (after i steps) l> 0), target value speed vr (i + l) (l> 0), and target value acceleration ar (i + l) (l> 0) can be calculated. Since the state variable xM (i), model input uM (i), target value r (i), target value speed vr (i), target value acceleration ar (i) of the mathematical model 14 at each step can be calculated. By using Expression (29), the model input change amount ⁇ uM (i + l) and l> 0 at each step can also be calculated.
- the model gain K1, the target value speed model gain Kvr1, and the target value acceleration model gain Kar1 in Expression (20) are given, the state variable xM (i) of the mathematical model 14 that does not break the constraint condition It is also possible to calculate the model input uM (i), the target value r (i), the target value speed vr (i), and the target value acceleration ar (i). As described above, a condition that does not break this control constraint is called a maximum output allowable set O ⁇ i.
- FIG. 5 is a diagram illustrating an example of the maximum output allowable set O ⁇ i of the motor control device 200 according to Embodiment 2 of the present invention.
- the vertical axis represents the state variable xM (i) of the mathematical model 14, and the horizontal axis represents the target value r (i).
- the inside of the polygon represents the maximum allowable output set O ⁇ i.
- the target value speed vr (i) and the target value acceleration ar (i) are also elements constituting the maximum output allowable set O ⁇ i, and should be added to the axis of FIG. Omitted for simplicity.
- Point A represents a situation where the state variable xM (i) and target value r (i) at the i step of the mathematical model 14 are within the maximum output allowable set O ⁇ i, and point B is at the i step.
- FIG. 6 is a diagram illustrating an operation example of the motor control device 200 according to Embodiment 2 of the present invention.
- the present embodiment is not limited to three sets of model gain, target value model gain, target value speed model gain, target value acceleration model gain, and the like.
- FIG. 6 shows three sets of model gain, target value model gain, target value speed model gain, target value acceleration model gain, ⁇ K1, Kr1, Kvr1, Kar1 ⁇ , ⁇ K2, Kr2, Kvr2, Kar2 ⁇ , ⁇ K3 in advance. , Kr3, Kvr3, Kar3 ⁇ are designed, and the maximum output allowable sets O ⁇ 1, O ⁇ 2, and O ⁇ 3 are calculated when the gains of each set are used.
- the vertical axis represents the state variable xM (i) of the mathematical model 14, and the horizontal axis represents the target value r (i).
- the gain that the control output y (i) follows the target value r (i) earliest is ⁇ K1, Kr1, Kvr1, Kar1 ⁇
- the gain that follows the latest is ⁇ K3, Kr3, Kvr3.
- Kar3 ⁇ The gain ⁇ K2, Kr2, Kvr2, Kar2 ⁇ is slower than the gain ⁇ K1, Kr1, Kvr1, Kar1 ⁇ and earlier than ⁇ K3, Kr3, Kvr3, Kar3 ⁇
- the control output y (i) is the target value r (i ). Since the maximum output allowable set O ⁇ i depends on the gain ⁇ Ki, Kri, Kvri, Kavri ⁇ , the maximum output allowable sets O ⁇ 1, O ⁇ 2, and O ⁇ 3 when the respective gains are used are shown in FIG. Different as shown.
- the state variable xM (i) and the target value r (i) of the mathematical model 14 are at the point A. Since the point A is inside the maximum output allowable set O ⁇ 3 and outside the maximum output allowable sets O ⁇ 1 and O ⁇ 2, when the gain ⁇ K3, Kr3, Kvr3, Kar3 ⁇ is used, the equation (20 ) Is not violated, but if the gain ⁇ K2, Kr2, Kvr2, Kar2 ⁇ , ⁇ K1, Kr1, Kvr1, Kar1 ⁇ is used, the control constraint of Expression (20) will be violated at some point. Therefore, the model controller determining unit 11A determines gains ⁇ K3, Kr3, Kvr3, Kar3 ⁇ that do not break the constraint of the equation (20), and outputs these selected gains to the model controller 12.
- the model controller determination unit 11A determines the gain that the control output y (i) follows the target value r (i) earliest. To do. That is, in this case, since the gain ⁇ K2, Kr2, Kvr2, Kar2 ⁇ is faster than the gain ⁇ K3, Kr3, Kvr3, Kar3 ⁇ , the control output y (i) follows the target value r (i).
- the unit determining unit 11A determines gains ⁇ K2, Kr2, Kvr2, Kar2 ⁇ and outputs them to the model controller 12.
- the model controller determination unit 11A determines the gain ⁇ K1, Kr1, Kvr1, Kar1 ⁇ that the control output y (i) follows the target value r (i) earliest among them, and the model controller 12 Output to.
- model controller determination unit 11A performs the following three operations.
- the target value vector rvec (i), the state variable xM (i) of the mathematical model 14 and the model input uM (i ⁇ 1) one step before are acquired, and the current values are the respective gains ⁇ Ki, Kri, Kvri, Kari ⁇ . To determine whether it is within or outside the maximum allowable output set O ⁇ i created. That is, it is determined whether or not the control constraint is violated when each gain ⁇ Ki, Kri, Kvri, Kari ⁇ is used.
- the gain ⁇ Ki, Kri, Kvri, Kari ⁇ that does not break the control constraint is determined and output to the model controller 12. If there are a plurality of gains ⁇ Ki, Kri, Kvri, Kari ⁇ that do not break the control constraint, the gain that the control output y (i) follows the target value r (i) earliest is determined.
- the model sub-controller 121 uses the model gain Ki, the target value model gain Kri, the target value speed model gain Kvri, and the target value acceleration model gain Kari output from the model controller determination unit 11A, using the equation (14). Is used to calculate the model input change amount ⁇ uM (i) and output it to the model input change amount adder 122.
- the model offset Gi is assumed to be 0 in the equation (14).
- the model input change amount adder 122 adds the one-step previous model input uM (i ⁇ 1) and the model input change amount ⁇ uM (i) using Expression (17), and adds the sum to the model input uM (i). Are output to the model input memory 13, the mathematical model 14, and the model input adder 3.
- the mathematical model 14 calculates the model output yM (i) and the state variable xM (i + 1) of the next step by using the equation (2) for the state variable xM (i) and the model input uM (i).
- the state variable xM (i + 1) of the next step is output to the model controller determination unit 11A and the model controller 12, and the model output yM (i) is output to the feedback control unit 2.
- the model input memory 13 stores the model input uM (i) for one step.
- the normative model unit 1A generates the model output yM (i) and the model input uM (i).
- the model input uM (i) is generated using the model input M (i ⁇ 1) one step before, as in the first embodiment.
- the model input change amount ⁇ uM (i) can be used as a new input.
- model input uM (i) Even if the model gain, the target value model gain, the target value speed model gain, and the target value acceleration model gain are switched, it is possible to prevent the model input uM (i) from changing rapidly. That is, by determining the model controller so that the model input becomes smaller than a predetermined value, the occurrence of torque saturation can be prevented and the model controller can be determined so that the amount of change in the model input becomes smaller than the predetermined value. By performing the above, it is possible to prevent the occurrence of voltage saturation and reduce the shock when the model controller is switched.
- the model input uM (i) is calculated by determining the model gain Ki, the target value model gain Kri, the target value speed model gain Kvri, the target value acceleration model gain Kari,
- the state variable xM (i), the target value r (i), the target value speed vr (i), the target value acceleration ar (i) is simply multiplied by each gain. Accordingly, since the calculation time can be shortened, the sampling time can also be shortened.
- the target value storage unit 15 By the target value storage unit 15, the target value speed vr (i) and the target value acceleration ar (i) Can be taken into account when calculating the model input change amount ⁇ uM (i), and a control system in which the control output y (i) follows the target value r (i) earlier than when these values are not used can be designed. .
- the model input one step before is used as the past model input.
- the model input is not necessarily limited to the model input one step before. It is easy to give the same effect by using the previous model input, and the same effect as this embodiment can be obtained even if a past model input other than the model input of one step before is used as the past model input. can get.
- FIG. 3 A motor control apparatus 300 according to Embodiment 3 of the present invention will be described with reference to FIGS.
- the model input uM (i ⁇ 1) one step before is used as the past model input uM ′ (i), and the target value memory 151 can store and hold the target value between two steps.
- this is limited to using the model input of the previous step as the past model input, or having the target value memory 151 provided with a number that can store and hold the target value for only two steps. It is not a thing.
- the model controller is determined using the maximum output allowable set O ⁇ i as in the second embodiment.
- the difference from the second embodiment is that the second embodiment has a model gain Ki. While the target value model gain Kri, the target value speed model gain Kvri, and the target value acceleration model gain Kari are switched, the present embodiment is that the controller structure itself is switched.
- the motor control device 300 includes a target value r (i) for the control output y (i) detected by the detector 5 representing the position and speed of the machine to be driven or the position and speed of the motor, and The control output y (i) is input, and the control input u (i) such as motor torque and current is output to the controlled object 4 so that the control output y (i) follows the target value r (i).
- the motor control device 300 includes a reference model unit 1 ⁇ / b> B, a feedback control unit 2, and a model input adder 3, and rotates the mechanical load and the mechanical load.
- a control object 4 composed of a motor such as a mold motor or a linear motor is connected to a detector 5 such as an encoder.
- the target value storage unit 15, the feedback control unit 2, the model input adder 3, the controlled object 4, and the detector 5 have the same configurations as those of the first and second embodiments, and thus description thereof is omitted.
- the target value r (i) is input to the reference model unit 1B, and the model output yM (i) representing the ideal operation of the control target 4 is output to the feedback control unit 2 and is output from the detector 5.
- a model input uM (i) for driving the controlled object 4 is output to the model input adder 3 so that the output y (i) follows the target value r (i).
- the normative model unit 1B stores and holds the target value r (i) for a predetermined time, together with the current target value r (i) to which one or more past target values stored and held are input.
- a target value storage unit 15 that outputs the target value vector rvec (i), a model controller determination unit 11B that determines the model controller 12 from a plurality of predetermined candidates, and a model output yM (i) that is the target value vector a model controller 12 that follows the target value r (i) included in rvec (i), a model input memory 13 that stores the model input uM (i), and a mathematical model 14 that simulates the characteristics of the controlled object 4;
- the operations of the model controller 12, the model input memory 13, the mathematical expression model 14, and the target value storage unit 15 are the same as those in the first and second embodiments, and thus the description thereof is omitted.
- the difference between the present embodiment and the motor control devices 100 and 200 of the first and second embodiments is the operation of the model controller determining unit 11B.
- a plurality of candidates designed in advance by pole arrangement or the like are selected.
- a method of determining by the model controller determination unit 11B will be described.
- the equation of state A case will be described in which three model sub-controllers 121C, 121D, and 121E represented in FIGS. 8 to 10 are designed in advance.
- the present embodiment is not limited to the use of model sub-controllers 121C, 121D, and 121E as model sub-controllers.
- the model sub-controller 121C in FIG. 8 sets the target value r (i), the target value r (i-1) one step before, the target value r (i-2) two steps before, the state variable xM ( i) and the previous model input uM (i ⁇ 1) are input, and the model input variation ⁇ uM (i) is output to the model input variation adder 122.
- the model sub-controller 121C receives the target value r (i) and the state variable xM (i) and outputs a model deviation eM (i) that is a difference between them, and a model subtractor 1215C.
- the target value speed vr (i) output from the calculator 1216C is input, the value is multiplied by the target value speed model gain Kvri, and the product is output to the adder 1218C. It includes a multiplication unit 1212C, the.
- the model sub-controller 121C inputs the target value r (i), the target value r (i-1) one step before, and the target value r (i-2) two steps before, and formula (16) is obtained.
- the target value acceleration ar (i) is calculated using the target value acceleration generator 1217C to be output to the target value acceleration model gain multiplier 1213C and the target value acceleration ar (i) output from the target value acceleration generator 1217C. This value is multiplied by the target value acceleration model gain Kari, and the product is output to the adder 1218C.
- the target value acceleration model gain multiplier 1213C and the model input uM (i ⁇ 1) one step before are input.
- a model input gain multiplier 1214C before calculating the product of the input value and the model input gain Kui one step before and outputting the calculation result to the adder 1218C. That.
- the model sub-controller 121C outputs the model deviation gain multiplier 1211C, the target value speed model gain multiplier 1212C, the target value acceleration model gain multiplier 1213C, and the one-step previous model input gain multiplier 1214C.
- An adder 1218C that adds all outputs and outputs the sum as a model input change amount ⁇ uM (i) is provided.
- the model sub-controller 121D in FIG. 9 sets the target value r (i), the target value r (i-1) one step before, the target value r (i-2) two steps before, the state variable xM ( i) and the previous model input uM (i ⁇ 1) are input, and the model input variation ⁇ uM (i) is output to the model input variation adder 122.
- the difference between the model sub-controller 121D of FIG. 9 and the model sub-controller 121C of FIG. 8 is that the integrated value memory 1212D is used to store the integrated value xc (i), and the stored integrated value xc (i ⁇ 1). Is added to the model deviation eM (i) to add an integration operation.
- the model sub-controller 121D includes a model deviation gain multiplier 1211C, a target value speed model gain multiplier 1212C, a target value acceleration model gain multiplier 1213C, and a one-step previous model input gain that are also used in the model sub-controller 121C.
- a multiplier 1214C, a model subtractor 1215C, a target value speed generator 1216C, and a target value acceleration generator 1217C are used, and description thereof will be omitted.
- the model sub-controller 121D adds the model deviation eM (i) and the one-step previous integrated value xc (i-1), and uses the sum as the integrated value xc (i) for the model deviation integral gain.
- the multiplier 1211D and the integrated value adder 1213D output to the integrated value memory 1212D and the integrated value xc (i) are input, stored for one sampling time, and the value stored after one sampling time is stored one step before the integrated value xc
- Model deviation integral gain multiplier 1211D, model deviation gain multiplier 1211C, target value speed model gain to be output to the calculator 1214D All the values output from the calculator 1212C, the target value acceleration model gain multiplier 1213C, the one-step previous model input gain multiplier 1214C, and the model deviation integral gain multiplier 1211D are added, and the sum is added to the model input change amount ⁇ uM (i ), An adder 1214D for outputting to the model input change amount adder 122 is added.
- the model sub-controller 121E in FIG. 10 sets the target value r (i), the target value r (i-1) one step before, the target value r (i-2) two steps before, the state variable xM ( i) and the previous model input uM (i ⁇ 1) are input, and the model input variation ⁇ uM (i) is output to the model input variation adder 122.
- the difference between the model sub-controller 121E in FIG. 10 and the model sub-controller 121C in FIG. 8 and the model sub-controller 121D in FIG. 9 is that a differential operation is added to the model sub-controller 121E by using the state variable memory 1212E. It is in the point.
- the model sub-controller 121E is a model deviation gain multiplier 1211C, a target value speed model gain multiplier 1212C, a target value acceleration model gain multiplier 1213C, which is also used in the model sub controller 121C, and is multiplied by a model input gain before one step. 1214C, a model subtractor 1215C, a target value speed generator 1216C, and a target value acceleration generator 1217C.
- model deviation integral gain multiplier 1211D an accumulated value memory 1212D, and an accumulated value adder 1213D, which are also used in the model controller 121D, are further provided. Since the description regarding these elements was mentioned above, it is abbreviate
- the model sub-controller 121E inputs the state variable xM (i), stores the input value for one sampling time, and after one sampling time, stores the stored value for the state variable xM one step before.
- the state variable memory 1212E output to the state variable differential subtractor 1213E as (i-1), the state variable xM (i-1) and the state variable xM (i) one step before output from the state variable memory 1212E are input.
- the state variable differential subtractor 1213E that subtracts the input value from the state variable xM (i) and outputs the difference to the state variable differential gain multiplier 1211E, and the value output from the state variable differential subtractor 1213E.
- An adder 1214E that adds all the values output from the differential gain multiplier 1211E and outputs the sum as a model input change amount ⁇ uM (i) to the model input change amount adder 122 is further provided.
- the model sub-controllers 121C, 121D, and 121E shown in FIGS. 8 to 10 use the model deviation gain Kei, the target value speed model gain Kvri, and the target value acceleration model.
- the gain Kari and the one-step previous model input gain Kui are represented by the same symbol, but these numerical values may be different for each model sub-controller 121C, 121D, 121E.
- the model deviation integral gain Kxii is also represented by the same symbol, but these numerical values may also be different in the model sub-controllers 121D and 121E.
- the model controller determination unit 11B determines whether the model input uM (i) One of a plurality of controller candidates designed in advance is determined so that the change amount ⁇ uM (i) does not exceed a predetermined value. By determining a plurality of controller candidates in advance, it is possible to reduce the amount of calculation on-line and to shorten the sampling time.
- the model controller determination unit 11B determines the model controller 12 using the maximum output allowable set O ⁇ i. In order to calculate the maximum allowable output set O ⁇ i, it is necessary to calculate the closed loop system of the reference model unit 1B as shown in the equation (30) in the second embodiment.
- a closed loop system is calculated when the model controller 12C of FIG. 8 is used.
- the model input change amount ⁇ uM (i) when the model sub-controller 121C of FIG. 8 is used is calculated as in the following equation (32).
- equation (32) the closed loop system when the model controller 12C is used is represented by the following equation (33).
- a closed loop system is calculated when the model controllers 12D and 12E shown in FIGS. 9 and 10 are used.
- the model input change amount ⁇ uM (i) output from the model sub-controller 121D is calculated by the following equations (34) and (35).
- the model input change amount ⁇ uM (i) output from the model sub-controller 121E is calculated by the following equation (37).
- Equation (38) a closed loop system using the model controller 12E is represented by the following Equation (38).
- Expressions (33), (36), and (38) are expressions that have the same form as Expression (30), the maximum output allowable set O ⁇ i is calculated by the same method as in the second embodiment. It is possible. Control in which the model controller determining unit 11B uses the maximum output allowable set O ⁇ i without breaking the control constraint equation (20) and the control output y (i) follows the target value r (i) earliest. If a device is selected, a control system can be obtained that does not violate the above-mentioned control constraints and has a fast follow-up. Since the method for determining the model controller 12 from a plurality of candidates predetermined by the model controller determination unit 11B using the maximum output allowable set O ⁇ i is the same as the method described in the second embodiment, here. Description is omitted.
- model input uM (i) is generated using model input uM (i ⁇ 1) one step before, as in the first and second embodiments.
- the model input change amount ⁇ uM (i) can be used as a new input.
- the model input uM (i) not only the model input uM (i) but also the control system in which the model input change amount ⁇ uM (i) does not become larger than a predetermined value can be easily realized. That is, it is possible to obtain a high-speed and high-precision motor control device while preventing torque saturation, torque change amount saturation, and voltage saturation closely related to torque change amount saturation. Even if the model controller determining unit determines a different model controller and calculates the model input uM (i), it is possible to prevent the model input uM (i) from changing rapidly.
- the calculation of the model input uM (i) is only the determination of the model controller and the calculation of the state variable xM (i) of the mathematical model 14 and the model controller. Therefore, since the calculation time can be shortened, the sampling time can also be shortened.
- the model input one step before is used as the past model input in the same manner as in the first and second embodiments.
- the model input is not necessarily limited to one step before. It is easy to give the same result by using the model input of, and even if a past model input other than the model input of one step before is used as the past model input, the same effect as this embodiment can be obtained. .
- the present invention is not limited to the above-described embodiment, and various modifications can be made without departing from the scope of the invention in the implementation stage.
- the above embodiments include inventions at various stages, and various inventions can be extracted by appropriately combining a plurality of disclosed constituent requirements. For example, even if some constituent elements are deleted from all the constituent elements shown in each of the first to third embodiments, the problem described in the column of the problem to be solved by the invention can be solved, and the column of the effect of the invention. When the effects described in (1) are obtained, a configuration in which this configuration requirement is deleted can be extracted as an invention. Furthermore, the structural requirements over the first to third embodiments may be combined as appropriate.
- the motor control device is useful for controlling a control target including a motor, and is particularly suitable for a motor control device that drives an industrial machine.
Landscapes
- Engineering & Computer Science (AREA)
- Power Engineering (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Automation & Control Theory (AREA)
- Control Of Electric Motors In General (AREA)
Abstract
Description
本発明の実施の形態1に係るモータ制御装置100について図1から図3までを参照しながら説明する。図1は、この発明の実施の形態1に係るモータ制御装置100の構成を示すブロック図である。なお、以降では、各図中、同一符号は同一又は相当する部分を示す。
本発明の実施の形態2に係るモータ制御装置200について図4から図6までを参照しながら説明する。なお、本実施の形態においても、説明の簡略化のため、過去モデル入力uM’(i)として1ステップ前のモデル入力uM(i-1)を用いる場合について説明する。また、目標値メモリ151は2ステップ間目標値を記憶保持することが可能な個数(つまりM=2)備えられている場合について説明する。ただし、これは過去モデル入力として1ステップ前のモデル入力を使用することや、目標値メモリ151が2ステップ間のみの目標値を記憶保持することが可能な個数備えられていることに限定されるものではない。
本発明の実施の形態3に係るモータ制御装置300について図7から図10までを参照しながら説明する。なお、本実施の形態においても過去モデル入力uM’(i)として1ステップ前のモデル入力uM(i-1)を、目標値メモリ151は2ステップ間目標値を記憶保持することが可能な個数(つまりM=2)備えられている場合について説明する。ただし、これは過去モデル入力として1ステップ前のモデル入力を使用することや、目標値メモリ151が2ステップ間のみの目標値を記憶保持することが可能な個数備えられていることに限定されるものではない。
2 フィードバック制御部
3 モデル入力加算器
4 制御対象
5 検出器
11、11A、11B モデル制御器決定部
12、12C、12D、12E モデル制御器
13 モデル入力メモリ
14 数式モデル
15 目標値記憶部
21 モデル出力減算器
22 フィードバック制御器
100、200、300 モータ制御装置
121、121C、121D、121E モデルサブ制御器
122 モデル入力変化量加算器
151 目標値メモリ
1211C モデル偏差ゲイン乗算器
1212C 目標値速度モデルゲイン乗算器
1213C 目標値加速度モデルゲイン乗算器
1214C 1ステップ前モデル入力ゲイン乗算器
1215C モデル減算器
1216C 目標値速度生成器
1217C 目標値加速度生成器
1218C、1214D、1214E 加算器
1211D モデル偏差積分ゲイン乗算器
1212D 積算値メモリ
1213D 積算値加算器
1211E 状態変数微分ゲイン乗算器
1212E 状態変数メモリ
1213E 状態変数微分減算器
Claims (12)
- モータを含む制御対象の制御出力を追従させるべき目標値に基づいて、前記制御対象の所望の動作を表すモデル出力と、前記所望の動作に前記制御対象を駆動するモデル入力とを生成する規範モデル部と、
前記制御出力と前記モデル出力を入力し、前記制御出力を前記モデル出力に追従させるフィードバック入力を生成するフィードバック制御部と、
前記モデル入力及び前記フィードバック入力を加算して前記制御対象への制御入力を生成するモデル入力加算器と、
を備えるモータ制御装置であって、
前記規範モデル部は、
前記目標値の現在の値および前記目標値の1つまたは複数の過去の値を目標値ベクトルとして保持する目標値記憶部と、
前記制御対象の特性を模擬し、前記モデル入力に基づいて前記モデル出力及び状態変数を生成する数式モデルと、
前記目標値ベクトル及び前記状態変数に基づいて、前記モデル入力を生成するモデル制御器と、
前記目標値ベクトル及び前記状態変数に基づいて、予め定めた複数のモデル制御器の候補から前記モデル制御器を決定するモデル制御器決定部と、
を含む、
ことを特徴とするモータ制御装置。 - 前記規範モデル部は、
前記モデル入力の過去の値である過去モデル入力を保持するモデル入力メモリをさらに備え、
前記モデル制御器は、前記過去モデル入力にも基づいて、前記モデル入力を生成し、
前記モデル制御器決定部は、前記過去モデル入力にも基づいて、前記モデル制御器を決定する
ことを特徴とする請求項1に記載のモータ制御装置。 - 前記モデル制御器は、前記目標値ベクトル、前記状態変数、及び前記過去モデル入力に基づいて前記モデル入力の変化量を計算し、前記過去モデル入力と前記変化量とを加算することにより前記モデル入力を生成する
ことを特徴とする請求項2に記載のモータ制御装置。 - 前記モデル制御器決定部は、前記目標値ベクトル、前記状態変数、及び前記過去モデル入力に基づいて数値ベクトルであるモデルゲインを決定し、
前記モデル制御器は、前記目標値ベクトル、前記状態変数、及び前記過去モデル入力に前記モデルゲインを乗じて前記変化量を生成する
ことを特徴とする請求項3に記載のモータ制御装置。 - 前記モデル制御器は、前記変化量にモデルオフセットを加算してから前記過去モデル入力と加算することにより、前記モデル入力を生成し、
前記モデル制御器決定部は、前記目標値ベクトル、前記状態変数、及び前記過去モデル入力に基づいて前記モデルオフセットを決定する
ことを特徴とする請求項4に記載のモータ制御装置。 - 前記モデル制御器決定部は、前記モデル入力及び前記変化量のどちらか一方、あるいは両方とも所定の値より小さくなるように前記モデル制御器の決定を行う
ことを特徴とする請求項3に記載のモータ制御装置。 - 前記モデル制御器決定部は、前記モデル入力及び前記変化量のどちらか一方、あるいは両方とも所定の値より小さくなるように前記モデルゲインを決定する
ことを特徴とする請求項4に記載のモータ制御装置。 - 前記モデル制御器決定部は、前記モデル入力及び前記モデル入力の変化量のどちらか一方、あるいは両方とも所定の値より小さくなるように前記モデルゲイン及び前記モデルオフセットを決定する
ことを特徴とする請求項5に記載のモータ制御装置。 - 前記モデル制御器決定部は、前記目標値ベクトル、前記状態変数、及び前記過去モデル入力に基づいて、予め定めた前記モデルゲインの候補を含んだゲインマップを用いて、前記モデルゲインを決定する
ことを特徴とする請求項4、5または7に記載のモータ制御装置。 - 前記モデル制御器決定部は、前記目標値ベクトル、前記状態変数、及び前記過去モデル入力に基づいて、予め定めた前記モデルゲインの候補を含んだゲインマップ及び予め定めた前記モデルオフセットの候補を含んだオフセットマップを用いて、前記モデルゲイン及びモデルオフセットを決定する
ことを特徴とする請求項5または8に記載のモータ制御装置。 - 前記数式モデルは、振動的な機械系をモデル化したものである
ことを特徴とする請求項1から11のいずれか1項に記載のモータ制御装置。
Priority Applications (3)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
US14/356,229 US9274514B2 (en) | 2011-11-22 | 2012-03-21 | Motor control apparatus |
JP2013545806A JP5726328B2 (ja) | 2011-11-22 | 2012-03-21 | モータ制御装置 |
DE112012004848.2T DE112012004848B9 (de) | 2011-11-22 | 2012-03-21 | Motorregelungsvorrichtung |
Applications Claiming Priority (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
JP2011-254929 | 2011-11-22 | ||
JP2011254929 | 2011-11-22 |
Publications (1)
Publication Number | Publication Date |
---|---|
WO2013077007A1 true WO2013077007A1 (ja) | 2013-05-30 |
Family
ID=48469470
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
PCT/JP2012/057217 WO2013077007A1 (ja) | 2011-11-22 | 2012-03-21 | モータ制御装置 |
Country Status (5)
Country | Link |
---|---|
US (1) | US9274514B2 (ja) |
JP (1) | JP5726328B2 (ja) |
DE (1) | DE112012004848B9 (ja) |
TW (1) | TWI440295B (ja) |
WO (1) | WO2013077007A1 (ja) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
EP3379349A1 (en) | 2017-03-14 | 2018-09-26 | Omron Corporation | Control device, control program, and control system |
Families Citing this family (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
TWI493303B (zh) * | 2013-10-23 | 2015-07-21 | Hurco Automation Ltd | Feedforward weight adjustment method and its calculation device |
TWI506943B (zh) * | 2013-12-05 | 2015-11-01 | Rdc Semiconductor Co Ltd | 運用於時變信號的信號處理裝置 |
CN104698970B (zh) * | 2013-12-06 | 2017-06-16 | 金丽科技股份有限公司 | 运用于时变信号的信号处理装置 |
US10044315B2 (en) * | 2016-07-15 | 2018-08-07 | Onesubsea Ip Uk Limited | Systems and methods for mitigating resonance in long cable drives |
CN107632519B (zh) * | 2016-07-18 | 2021-01-26 | 中国北方车辆研究所 | 一种双缸协同运动机构的稳定控制方法 |
EP3324254A1 (de) * | 2016-11-17 | 2018-05-23 | Siemens Aktiengesellschaft | Einrichtung und verfahren zur bestimmung der parameter einer regeleinrichtung |
RU2661749C1 (ru) * | 2017-09-11 | 2018-07-19 | федеральное государственное казенное военное образовательное учреждение высшего образования "Военная академия связи имени Маршала Советского Союза С.М. Буденного" Министерства обороны Российской Федерации | Устройство вычисления многопараметрического интегрального показателя |
Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2005285030A (ja) * | 2004-03-31 | 2005-10-13 | Yaskawa Electric Corp | サーボ制御装置 |
JP2010130852A (ja) * | 2008-11-28 | 2010-06-10 | Yaskawa Electric Corp | 電動機制御装置と予測制御方法 |
WO2011136160A1 (ja) * | 2010-04-26 | 2011-11-03 | 三菱電機株式会社 | サーボ制御装置 |
Family Cites Families (11)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CA2068329C (en) * | 1990-09-18 | 1998-10-13 | Ronald James Kadlec | Digital servo control system for use in disk drives |
JP2968631B2 (ja) * | 1992-01-31 | 1999-10-25 | 新日本製鐵株式会社 | 熱延プロセス制御方法 |
JPH08152902A (ja) * | 1994-11-30 | 1996-06-11 | Hitachi Ltd | 適応処理装置 |
JP3740189B2 (ja) * | 1995-07-18 | 2006-02-01 | 株式会社日立グローバルストレージテクノロジーズ | 位置決め適応制御装置並びに情報記憶装置 |
JPH0973618A (ja) * | 1995-09-07 | 1997-03-18 | Toshiba Corp | ディスク記録再生装置のヘッド位置決め制御システム及びそのシステムに適用する速度制御方法 |
US6545836B1 (en) * | 1999-11-12 | 2003-04-08 | Acorn Technologies, Inc. | Servo control apparatus and method using absolute value input signals |
JP3863069B2 (ja) * | 2002-06-06 | 2006-12-27 | 本田技研工業株式会社 | プラントの制御装置 |
WO2004057584A2 (en) * | 2002-12-19 | 2004-07-08 | Koninklijke Philips Electronics N.V. | Disc drive with improved resistance against mechanical shocks |
US8065022B2 (en) * | 2005-09-06 | 2011-11-22 | General Electric Company | Methods and systems for neural network modeling of turbine components |
JP2010073284A (ja) * | 2008-09-19 | 2010-04-02 | Hitachi Global Storage Technologies Netherlands Bv | ヘッドのサーボ制御システム及びヘッドのサーボ制御方法 |
JP5308799B2 (ja) * | 2008-12-04 | 2013-10-09 | エイチジーエスティーネザーランドビーブイ | ディスク・ドライブ及びそのヘッドのサーボ制御方法 |
-
2012
- 2012-03-21 DE DE112012004848.2T patent/DE112012004848B9/de not_active Expired - Fee Related
- 2012-03-21 WO PCT/JP2012/057217 patent/WO2013077007A1/ja active Application Filing
- 2012-03-21 JP JP2013545806A patent/JP5726328B2/ja active Active
- 2012-03-21 US US14/356,229 patent/US9274514B2/en not_active Expired - Fee Related
- 2012-04-06 TW TW101112199A patent/TWI440295B/zh active
Patent Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2005285030A (ja) * | 2004-03-31 | 2005-10-13 | Yaskawa Electric Corp | サーボ制御装置 |
JP2010130852A (ja) * | 2008-11-28 | 2010-06-10 | Yaskawa Electric Corp | 電動機制御装置と予測制御方法 |
WO2011136160A1 (ja) * | 2010-04-26 | 2011-11-03 | 三菱電機株式会社 | サーボ制御装置 |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
EP3379349A1 (en) | 2017-03-14 | 2018-09-26 | Omron Corporation | Control device, control program, and control system |
US10386794B2 (en) | 2017-03-14 | 2019-08-20 | Omron Corporation | Control device, storage medium, and control system by creating internal model of control target |
Also Published As
Publication number | Publication date |
---|---|
TW201322616A (zh) | 2013-06-01 |
DE112012004848T5 (de) | 2014-09-04 |
JPWO2013077007A1 (ja) | 2015-04-27 |
JP5726328B2 (ja) | 2015-05-27 |
TWI440295B (zh) | 2014-06-01 |
DE112012004848B9 (de) | 2018-03-22 |
DE112012004848B4 (de) | 2018-01-11 |
US9274514B2 (en) | 2016-03-01 |
US20140306641A1 (en) | 2014-10-16 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
JP5726328B2 (ja) | モータ制御装置 | |
JP5312688B2 (ja) | サーボ制御装置 | |
JP4541218B2 (ja) | 指令生成装置 | |
JP4697139B2 (ja) | サーボ制御装置 | |
JP2011501293A (ja) | 運動制御システムにおいて終了条件を満たす方法およびシステム | |
EP2846207B1 (en) | Sizing and tuning methodology for optimized motion control components and energy efficiency | |
JP2017102617A (ja) | 補正装置、補正装置の制御方法、情報処理プログラム、および記録媒体 | |
JP2009503708A (ja) | 機械における機械要素の移動案内のための方法および装置 | |
CN105027013A (zh) | 机床的进给轴控制方法及进给轴控制装置 | |
CN104204977A (zh) | 轨迹控制装置 | |
CN111033395B (zh) | 控制系统 | |
CN110941242A (zh) | 电动机控制装置 | |
JP5441944B2 (ja) | モータ制御装置 | |
Chu et al. | Efficient model predictive algorithms for tracking of periodic signals | |
JP2014117787A (ja) | 制御装置 | |
WO2002025390A1 (fr) | Procede de servocommande | |
Manna et al. | Comparative performance analysis of LQR and MPC for active suspension system | |
CN111487930B (zh) | 基于对称图形替换技术的运动控制系统、方法、装置、机床和存储介质 | |
Kia et al. | Modeling and simulation of a single gain tuning ADRC controller in Matlab/Simulink | |
TWI406116B (zh) | 數值控制方法及裝置 | |
JP5573664B2 (ja) | 数値制御装置、移動制御方法、移動制御プログラム及び記憶媒体 | |
CN111052001A (zh) | 控制装置 | |
Khasanov et al. | The modeling of multicircuit control systems in robotized technological complex | |
Ruderman et al. | Analysis of settling behavior and design of cascaded precise positioning control in presence of nonlinear friction | |
Nguyen et al. | Integrator-Backstepping Based Control for Nonlinear Roll-to-Roll Web Dynamics |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
121 | Ep: the epo has been informed by wipo that ep was designated in this application |
Ref document number: 12852217 Country of ref document: EP Kind code of ref document: A1 |
|
ENP | Entry into the national phase |
Ref document number: 2013545806 Country of ref document: JP Kind code of ref document: A |
|
WWE | Wipo information: entry into national phase |
Ref document number: 14356229 Country of ref document: US |
|
WWE | Wipo information: entry into national phase |
Ref document number: 1120120048482 Country of ref document: DE Ref document number: 112012004848 Country of ref document: DE |
|
122 | Ep: pct application non-entry in european phase |
Ref document number: 12852217 Country of ref document: EP Kind code of ref document: A1 |