JP5567382B2 - PID parameter adjustment support apparatus and method - Google Patents

Description

  The present invention relates to a PID parameter adjustment support apparatus and method for supporting adjustment of PID parameters of a PID controller.

When adjusting the PID parameter of the PID controller using the PID parameter adjustment support tool, the recommended PID parameter is obtained and displayed as a set based on an adjustment variation based on a single adjustment method or a single adjustment theory background. (See Patent Document 1).
In such a PID parameter adjustment support tool, a PID parameter is obtained as a set based on a predetermined single adjustment method or a single adjustment method selected by the adjustment support tool or the user. Examples of adjustment methods include the Ziegler-Nichols method and the CHR (Chien, Hrones, Reswick) method.

JP 2009-116515 A

  A parameter calculation function used in a PID parameter adjustment support tool or the like often employs a knowledge type adjustment method such as the ZN method or the CHR method, and can select a PID parameter among variations of these adjustment methods. It is like that.

  However, the variations of these adjustment methods are only variations corresponding to differences in the overshoot amount regulations, etc., and the theoretical background is similar, that is, based on a single adjustment theory background, so control When parameter adjustment as expected could not be performed due to the convenience of the characteristics of the target, there was a problem that adjustment as expected could not be performed for all options.

  The present invention has been made to solve the above problems, and an object of the present invention is to provide a PID parameter adjustment support apparatus and method that can reduce the probability that a user will get stuck in PID parameter adjustment.

The PID parameter adjustment support apparatus of the present invention includes an input information processing means for acquiring information necessary for searching for a PID parameter, a knowledge type adjusting means for executing a knowledge type adjustment method and searching for a recommended value of the PID parameter, An analysis type adjustment means for searching for a recommended value of a PID parameter by executing an analysis type adjustment method; an heuristic type adjustment means for searching for a recommended value of a PID parameter by executing an empirical rule type readjustment technique; Adjustment procedure control means for causing at least one of knowledge type adjustment means, analysis type adjustment means, and empirical rule type adjustment means to execute processing, knowledge type adjustment means, analysis type adjustment means, and empirical rule type adjustment whenever executing the processing by any means, and an information presenting means for presenting the recommended value of the PID parameters obtained in this process, the input information processing means, the knowledge-based A simple modeling unit that generates a control target model used by the adjustment unit using a simple modeling method, and a control target model used by the analytical adjustment unit is generated from a control result based on the adjustment of the knowledge type adjustment unit. A control action base modeling unit; and a control task information input unit for inputting a control task used by the rule-of-thumb adjustment unit. The adjustment procedure control unit includes the knowledge type adjustment unit according to a user instruction, At least one means is made to execute processing in the order of the analysis type adjustment means and the heuristic type adjustment means .

The PID parameter adjustment support apparatus of the present invention includes an input information processing means for acquiring information necessary for searching for a PID parameter, and a knowledge type adjusting means for searching for a recommended value of the PID parameter by executing a knowledge type adjustment technique. And an analysis type adjustment unit that searches for a recommended value of a PID parameter by executing an analysis type adjustment method, and an empirical rule type adjustment unit that searches for a recommended value of a PID parameter by executing a rule of thumb type readjustment method , Adjustment procedure control means for causing at least one of the knowledge type adjustment means, the analysis type adjustment means, and the empirical rule type adjustment means to perform processing, the knowledge type adjustment means, the analysis type adjustment means, the empirical rule Each time the processing by one of the type adjustment means is executed, information presentation means for presenting a recommended value of the PID parameter obtained by this processing, the knowledge type adjustment means, and the analysis type adjustment Stage, and a continuation determination means for determining whether to continue the search by the empirical type adjustment means, said input information processing means, a simple modeling approach the controlled object model which the knowledge-based adjustment means is utilized A simple modeling means for generating using, a control action based modeling means for generating a control target model used by the analysis type adjusting means from a control result based on the adjustment of the knowledge type adjusting means, and the heuristic type adjusting means. A control task information input unit that receives a control task to be used from a user, and the continuation determination unit includes a feature value obtained from a result of controlling a control target by a PID controller in which a recommended value of the PID parameter is set, and a user Whether to continue searching for the recommended value of the PID parameter by comparing with the control target set by the knowledge type adjustment And determining each time the processing of the analysis type adjustment unit is completed, and the adjustment procedure control unit determines whether the knowledge type adjustment unit, the analysis type adjustment unit, and the empirical rule type according to the determination of the continuation determination unit At least one means is made to execute processing in the order of the adjustment means .

Further, in one configuration example of the PID parameter adjustment support device of the present invention, the information presentation unit further executes a simulation for controlling a controlled object by a PID controller in which a recommended value of the PID parameter is set, and a result of the simulation The obtained control response is presented.
In the configuration example of the PID parameter adjustment support apparatus according to the present invention, the information presentation unit further presents a description of the type of each adjustment method.

  In one configuration example of the PID parameter adjustment support apparatus of the present invention, the knowledge type adjustment method searches for a recommended value of the PID parameter by any one of the Ziegler-Nichols method, the CHR method, the IMC method, and the Cohen Coon method. The analytical adjustment method calculates an evaluation function value indicating the proximity between a control response obtained by simulation and a predetermined ideal control response characteristic, and sequentially changes PID parameters. The simulation is repeatedly executed to search for a recommended value of the PID parameter at which the evaluation function value is an optimum value. The rule-of-thumb re-adjustment method is used to control problems generated in the control response obtained by the simulation. The PID based on a readjustment rule set in advance according to the selected control task. It is characterized in that is to explore the recommended value of the parameter.

The PID parameter adjustment support method of the present invention includes an input information processing step for acquiring information necessary for searching for a PID parameter, a knowledge type adjustment step for executing a knowledge type adjustment method to search for a recommended value of the PID parameter, and analysis At least one of an analysis type adjustment step for searching for a recommended value of a PID parameter by executing a type adjustment method, and an empirical type adjustment step for searching for a recommended value of a PID parameter by executing a rule of thumb type readjustment method The recommended value of the PID parameter obtained in this process each time the process according to any one of the adjustment procedure control step, the knowledge type adjustment step, the analysis type adjustment step, and the heuristic type adjustment step is executed. and a information presenting step of presenting, the input information processing step, mode of the control target of the knowledge-based adjustment step utilizes A simple modeling step for generating a model using a simple modeling method, a control action-based modeling step for generating a control target model used by the analysis type adjustment step from a control result based on the adjustment of the knowledge type adjustment step, the empirical rule At least one of control task information input steps for inputting a control task used by the mold adjustment step is performed, and the adjustment procedure control step includes the knowledge type adjustment step, the analysis type adjustment step according to a user instruction, At least one step is executed in the order of the rule-of-thumb adjustment step .

The PID parameter adjustment support method of the present invention includes an input information processing step for acquiring information necessary for searching for a PID parameter, and a knowledge type adjustment step for searching for a recommended value of the PID parameter by executing a knowledge type adjustment method. At least of an analysis type adjustment step of searching for a recommended value of a PID parameter by executing an analysis type adjustment method, and an empirical rule type of adjusting step of searching for a recommended value of a PID parameter by executing an empirical rule type readjustment method Whenever the process according to any one of the adjustment procedure control step, the knowledge type adjustment step, the analysis type adjustment step, and the heuristic type adjustment step is executed, the PID parameter obtained by this process is recommended. An information presentation step for presenting a value, the knowledge type adjustment step, the analysis type adjustment step, and the rule of thumb type adjustment step According and a continuation determination step of determining whether to continue searching, the input information processing step, simplified modeling step of generating with a simple modeling techniques the controlled object model which the knowledge-based adjustment step is utilized, Control action-based modeling step for generating a control target model used by the analysis type adjustment step from a control result based on the adjustment of the knowledge type adjustment step, control issue information for inputting a control issue used by the empirical rule type adjustment step At least one of the input steps is executed, and the continuation determining step includes a feature value obtained from a result of controlling a control target by a PID controller in which a recommended value of the PID parameter is set and a control target set by a user. Whether to continue searching for the recommended value of the PID parameter by comparing Is determined each time the processes of the knowledge type adjustment step and the analysis type adjustment step are completed, and the adjustment procedure control step determines the knowledge type adjustment step and the analysis according to the determination result of the continuation determination step. At least one step is executed in the order of the mold adjustment step and the heuristic type adjustment step .

  According to the present invention, the user recognizes by presenting the recommended values of the PID parameters by all methods of the knowledge type adjustment method, the analysis type adjustment method, the heuristic type readjustment method, or at least two types of methods. The user can appropriately select an appropriate recommended PID parameter value according to the accuracy of modeling to be performed and the scene of PID parameter adjustment such as initial adjustment or readjustment. As a result, in the present invention, it is possible to reduce the probability that the user will get stuck in the PID parameter adjustment.

  In addition, in the present invention, whenever a process using any one of the knowledge type adjustment method, the analysis type adjustment method, and the heuristic type readjustment method is executed, the recommended value of the PID parameter obtained by this process is presented. Therefore, it is possible to make the user determine whether or not to continue searching for the recommended value of the PID parameter, and it is possible to further reduce the probability that the user will get stuck in the PID parameter adjustment.

  In the present invention, the user's input work can be reduced by providing a continuation determination unit that determines whether or not to continue the search by the knowledge type adjustment unit, the analysis type adjustment unit, and the heuristic type adjustment unit. .

  Further, in the present invention, a knowledge type adjustment method, an analysis type is performed by executing a simulation for controlling a controlled object by a PID controller in which a recommended value of a PID parameter is set and presenting a control response obtained as a result of the simulation. It is possible to make the user confirm the control response predicted by the PID parameter according to each of the adjustment methods and the heuristic-type readjustment method before executing the control operation.

  Further, in the present invention, by presenting an explanation of the type of each adjustment method, the user can recognize the characteristics of each adjustment method.

It is a flowchart which shows the flow of a process of the PID parameter adjustment assistance method which concerns on the 1st Embodiment of this invention. It is a block diagram which shows the structure of the PID parameter adjustment assistance apparatus which concerns on the 1st Embodiment of this invention. It is a figure which shows an example of the adjustment information presentation screen in the 1st Embodiment of this invention. It is a flowchart which shows the flow of a process of the PID parameter adjustment assistance method which concerns on the 2nd Embodiment of this invention. It is a block diagram which shows the structure of the PID parameter adjustment assistance apparatus which concerns on the 2nd Embodiment of this invention. It is a figure which shows an example of the knowledge type | mold adjustment method selection screen in the 2nd Embodiment of this invention. It is a figure which shows an example of the adjustment information presentation screen in the 2nd Embodiment of this invention. It is a figure which shows an example of the adjustment information presentation screen in the 2nd Embodiment of this invention. It is a flowchart which shows the flow of a process of the PID parameter adjustment assistance method which concerns on the 3rd Embodiment of this invention. It is a block diagram which shows the structure of the PID parameter adjustment assistance apparatus which concerns on the 3rd Embodiment of this invention. It is a figure which shows an example of the control target setting screen in the 3rd Embodiment of this invention.

[Principle of Invention 1]
Different types of adjustments such as the heuristic-type readjustment method with a user interface disclosed in JP 2000-148204 A and the analysis type adjustment method disclosed in JP 2005-070940 A It was noticed that the probability of reaching a deadlock state of PID parameter adjustment can be reduced by providing a backup method and adding a function for guiding the user to utilize these methods. Then, I came up with the idea of inducing the proper use as follows.

  For example, the accuracy of modeling recognized by the user by presenting the recommended values of PID parameters by all methods of knowledge type adjustment method, rule of thumb type readjustment method, analysis type adjustment method or any two methods In addition, the user can appropriately select an appropriate recommended PID parameter value according to the scene of PID parameter adjustment such as initial adjustment or readjustment. It is more preferable to provide a description of the type of each adjustment method. As the modeling accuracy recognized by the user, for example, relatively high-precision modeling such as modeling of first-order lag approximation of the operation amount MV step input, time series data matching utilizing the operation amount MV of the control operation by changing the set value SP. and so on.

[Principle of Invention 2]
By inducing the appropriate use of the following, we have come up with a solution to the problem.
(A) For example, if priority is given to the safety aspect of control, a simple and surely safe response waveform is obtained by step input of the manipulated variable MV, and a general knowledge type PID is obtained by approximating the controlled object with a first order delay. First, the user performs parameter adjustment. The response waveform by step input of the manipulated variable MV is different from the operation in the case of performing the PID control, and the response range in which the manipulated variable MV or the controlled variable PV has a fluctuation range and the duration of the fluctuation cannot be unexpectedly increased. It is a waveform. Then, the user confirms the response by the PID control after adjusting the knowledge type PID parameter.

  Even if the response waveform by this PID control is inappropriate, the response is usually in a safe range. That is, since the balance of P, I, and D is in a range that is determined as common sense, the response waveform is such that an abnormal PID control operation that cannot be used for analysis does not occur. Accordingly, the response waveform data obtained by the PID control at this time is useful data for performing the analysis type PID parameter adjustment, and the analysis type PID parameter adjustment is subsequently performed by the user. Then, the user confirms the response by the PID control after adjusting the analysis type PID parameter.

  Even if the response waveform by this PID control is inappropriate, it usually converges to a limited range of problems such as overshoot and hunting. Therefore, heuristic rule type PID parameter readjustment is continued.

(B) On the other hand, instead of giving priority to the safety aspect of the control as described above, for example, if priority is given to the reduction of adjustment labor, only the feature point of the control amount PV from the control response by actually executing PID control. By using a simple modeling method disclosed in Japanese Patent Application Laid-Open No. 2006-127079, for example, the control object is first-order delayed and a general knowledge type PID parameter adjustment is first performed by the user. Then, the user confirms the response by the PID control after adjusting the knowledge type PID parameter. If a good control result is obtained by this adjustment, a minimum amount of work is required.

  Even if the response waveform by PID control is inappropriate, the response is usually in a safe range. Accordingly, the response waveform data obtained by the PID control at this time is useful data for performing the analysis type PID parameter adjustment, and the analysis type PID parameter adjustment is subsequently performed by the user. Then, the user confirms the response by the PID control after adjusting the analysis type PID parameter.

  Even if the response waveform by this PID control is inappropriate, it usually converges to a limited range of problems such as overshoot and hunting. Therefore, heuristic rule type PID parameter readjustment is continued.

[First Embodiment]
Hereinafter, embodiments of the present invention will be described with reference to the drawings. This embodiment corresponds to Principle 1 of the invention described above. FIG. 1 is a flowchart showing the flow of processing of a PID parameter adjustment support method according to the first embodiment of the present invention, and FIG. 2 shows the configuration of the PID parameter adjustment support apparatus according to the first embodiment of the present invention. It is a block diagram.

  The PID parameter adjustment support apparatus includes an input information processing unit 1 that acquires information necessary for searching for a PID parameter, a control parameter adjustment unit 2 that searches for a PID parameter, and a recommended value of the PID parameter searched by the control parameter adjustment unit 2 It is comprised from the adjustment information presentation part 3 which presents.

  The input information processing unit 1 includes a modeling unit 10 that generates a model to be controlled using a simple modeling method. The control parameter adjustment unit 2 executes a knowledge type adjustment method to search for a recommended value of the PID parameter, and an analysis type to execute the analysis type adjustment method to search for a recommended value of the PID parameter. And an adjustment unit 21.

  Hereinafter, the operation of the PID parameter adjustment support apparatus according to the present embodiment will be described with reference to FIG. First, the modeling unit 10 generates a model to be controlled (step S1-1 in FIG. 1). Here, as a modeling technique used by the modeling unit 10, a simple modeling technique disclosed in Japanese Patent Application Laid-Open No. 2006-127079 is used.

  The simple modeling method disclosed in Japanese Patent Application Laid-Open No. 2006-127079 is to generate a control target model by searching for the value of a control target model parameter of a mathematical model to be controlled. A control system model is virtually generated by combining a controller algorithm for generation and a mathematical model to be controlled, and a simulation for simulating a control response is executed for this control system, and a model control feature amount of the control system model (for example, Target value arrival time, overshoot amount, hunting cycle, etc.) and evaluation showing the proximity of the known actual control feature value obtained from the actual control target control result and the corresponding model control feature value The function value is calculated and the controlled model parameter value set in the controlled model is changed sequentially. Repeatedly executing the simulation, in which so as to determine the value of the control object model parameter evaluation function value is the optimum value (0 or minimum value close to 0) as a model parameter search result.

  Examples of the value of the control target model parameter include a gain Kp, a dead time Lp, and a time constant T1. In Japanese Patent Laid-Open No. 2006-127079, the control target is expressed by a first-order lag transfer function as having a first-order lag and dead time elements, but the control target is not limited to the first-order lag. .

  Next, the knowledge type adjustment unit 20 of the control parameter adjustment unit 2 searches for the PID parameter by using the knowledge type adjustment method disclosed in Japanese Patent Application Laid-Open No. 2009-116515 for the control target model generated by the modeling unit 10. Then, the adjustment information presentation unit 3 presents the PID parameter searched by the knowledge type adjustment unit 20 to the user (step S1-2 in FIG. 1). At this time, the adjustment information presenting unit 3 presents the user with information that the PID parameter is obtained by the knowledge type adjustment method. Knowledge type adjustment methods include the Ziegler-Nichols method, the CHR method, the IMC (Internal Model Control) method, and the Cohen Coon method. As is well known, the PID parameter includes a set of a proportional band Pb, an integration time Ti, and a differentiation time Td.

  Subsequently, the analysis type adjustment unit 21 of the control parameter adjustment unit 2 searches for the PID parameter by using the analysis type adjustment method disclosed in Japanese Patent Laid-Open No. 2005-070940 for the control target model generated by the modeling unit 10. Then, the adjustment information presentation unit 3 presents the PID parameters searched by the analysis type adjustment unit 21 to the user (step S1-3 in FIG. 1). At this time, the adjustment information presenting unit 3 presents the user with information that the PID parameter is obtained by the analytical adjustment method. Moreover, it is preferable that the screen on which information is presented is the same as the screen on which the information in step S1-2 is presented.

  The analysis type adjustment method disclosed in Japanese Patent Application Laid-Open No. 2005-070940 is based on a combination of a controller algorithm for virtually generating a PID controller and a control target model generated by the modeling unit 10 to virtually control a control system model. A simulation that simulates the control response for this control system is executed based on the controller operation constraints (for example, the upper and lower limits of the operation amount), and the control response obtained by the simulation and a predetermined ideal The evaluation function value indicating the closeness to the control response characteristic is calculated, and the simulation is repeatedly executed while sequentially changing the PID parameter of the controller algorithm, so that the evaluation function value is the optimum value (0 or the minimum value close to 0). The PID parameter that becomes is fixed as the PID parameter search result Than is.

  In this embodiment, the combination of the knowledge type adjustment method and the analysis type adjustment method has been described as an example. However, the present invention is not limited to this, and the analysis type adjustment method and Japanese Patent Application Laid-Open No. 2000-148204 are described. May be combined with the rule-of-thumb re-adjustment method disclosed in the above, or the knowledge-type adjustment method and the rule-of-thumb re-adjustment method may be combined. A heuristic readjustment method may be combined.

  When the rule-of-thumb readjustment method is used, the rule-of-thumb adjustment unit is provided in the control parameter adjustment unit 2. The rule-of-thumb adjustment unit searches for PID parameters by the rule-of-thumb re-adjustment method disclosed in Japanese Patent Application Laid-Open No. 2000-148204, and the adjustment information presentation unit 3 sets the PID parameters searched by the rule-of-thumb adjustment unit. Present to the user.

  The rule-of-thumb readjustment method disclosed in Japanese Patent Laid-Open No. 2000-148204 prepares in advance a trouble pattern (for example, a situation where hunting occurs or a situation where the response is too slow) caused by an actual control response. The user is allowed to select from the representative trouble patterns, and the PID parameter readjustment value of the controller algorithm based on the PID parameter readjustment rule set in advance based on the empirical rule according to the selected trouble pattern Is calculated.

  In the present embodiment, it is more preferable to present the explanation of the type of each adjustment method as information relating to the PID parameter adjustment. Specifically, for the knowledge type adjustment method, information such as “knowledge type characteristics: particularly recommended for standard initial adjustment based on a simple characteristic grasp of the control target” is presented. For the analysis type adjustment method, information such as “analysis type characteristics: particularly recommended for adjustment specific to the control object based on precise characteristic grasp of the control object” is presented. For the rule-of-thumb re-adjustment method, information such as “feature rule-type features: particularly recommended for re-adjustment for the purpose of improving control operation defects” is presented.

  FIG. 3 shows an example of an adjustment information presentation screen on which the adjustment information presentation unit 3 presents information. In the adjustment information presentation screen 300 shown in FIG. 3, an adjustment result (PID parameter adjustment value) 301 by the knowledge type adjustment unit 20 and an adjustment result 302 by the analysis type adjustment unit 21 are presented. A description 303 of the knowledge type adjustment method is presented together with the adjustment result 301, and a description 304 of the analysis type adjustment method is presented together with the adjustment result 302.

  In addition, on the adjustment information presentation screen 300, the explanation 305 of the modeling method of the modeling unit 10 is presented as model information, and the transfer function expression information 306 of the control target model generated by the modeling unit 10 is presented. Further, the adjustment information presentation screen 300 includes a control response obtained as a result of executing a simulation for controlling the control target model generated by the modeling unit 10 by the PID controller in which the knowledge type PID parameter adjustment by the knowledge type adjustment unit 20 is performed. A certain knowledge type adjustment prediction response, and an analysis type adjustment prediction response which is a control response obtained as a result of executing a simulation of controlling the control target model by the PID controller which has adjusted the analysis type PID parameter by the analysis type adjustment unit 21; An adjustment result graph 307 representing a model response that reproduces an actual control operation before adjustment is presented. Calculation of the knowledge type adjustment prediction response, the analysis type adjustment prediction response, and the model response is performed by the adjustment information presentation unit 3.

  As described above, in the present embodiment, the recommended values of the PID parameters are presented by all methods of the knowledge type adjustment method, the analysis type adjustment method, the heuristic type readjustment method, or at least two types of methods. Thus, the user can appropriately select an appropriate recommended PID parameter value according to the modeling accuracy recognized by the user and the scene of PID parameter adjustment such as initial adjustment or readjustment. As a result, in the present embodiment, it is possible to reduce the probability that the user will get stuck in the PID parameter adjustment.

  As a modeling technique used by the modeling unit 10, a modeling technique disclosed in Japanese Patent Application Laid-Open No. 2004-38428 may be used. The modeling technique disclosed in Japanese Patent Application Laid-Open No. 2004-38428 provides time series data of an operation amount MV to a mathematical model to be controlled, and time series data of a control amount PV output from the model in accordance with the operation amount MV. The control amount PV time series data collected from the actual control object and the control amount PV time series data output from the model are used to control using the optimization method such as the Powell method. The value of the control target model parameter of the target mathematical model is determined.

[Second Embodiment]
Next, a second embodiment of the present invention will be described. This embodiment corresponds to Principle 2 of the invention described above. FIG. 4 is a flowchart showing the flow of processing of the PID parameter adjustment support method according to the second embodiment of the present invention, and FIG. 5 shows the configuration of the PID parameter adjustment support apparatus according to the second embodiment of the present invention. It is a block diagram.

  The PID parameter adjustment support apparatus according to the present embodiment includes an input information processing unit 1a, a control parameter adjustment unit 2a, an adjustment information presentation unit 3a, a knowledge type adjustment unit, an analysis type adjustment unit, and an empirical rule type adjustment unit described later. It is comprised from the adjustment procedure control part 4 which makes at least one perform a process.

  The input information processing unit 1a has a simple modeling unit 11 that generates a control target model used by a knowledge type adjustment unit described later using a simple modeling method, and a control target model used by an analysis type adjustment unit described later. It has a control action base modeling unit 12 generated from a control result based on the adjustment of the mold adjusting unit, and a control task information input unit 13 for inputting a control task used by an empirical rule type adjusting unit described later. The control parameter adjustment unit 2a executes a knowledge type adjustment method to search for a recommended value of the PID parameter, and an analysis type to execute the analysis type adjustment method to search for a recommended value of the PID parameter. An adjustment unit 21a and an empirical rule type adjustment unit 22 that searches for a recommended value of the PID parameter by executing an empirical rule type readjustment method.

  Hereinafter, the operation of the PID parameter adjustment support apparatus according to the present embodiment will be described with reference to FIG. First, the adjustment procedure control unit 4 causes the simple modeling unit 11 to perform modeling. The simple modeling unit 11 generates a model to be controlled by a simple modeling method disclosed in Japanese Patent Application Laid-Open No. 2006-127079 (step S4-1 in FIG. 4). In Japanese Patent Laid-Open No. 2006-127079, the control target is expressed by a first-order lag transfer function as having a first-order lag and dead time elements, but the control target is not limited to the first-order lag. .

  Subsequently, the adjustment procedure control unit 4 assumes that the data collection process for modeling is simple, and the PID parameter set is not special (the ratio of P, I, and D is fixed based on past expertise). First, the knowledge type adjustment technique is selected by the user (step S4-2 in FIG. 4). FIG. 6 shows an example of the knowledge type adjustment method selection screen 600 in which the adjustment procedure control unit 4 presents the options of the knowledge type adjustment method to the user. As shown in FIG. 6, there are Ziegler-Nichols method, CHR method, IMC method, and Cohen Coon method as options of the knowledge type adjustment method.

  The fact that the data collection process for modeling is simple means that modeling is performed with relatively limited information, and the identification of process gain, process time constant, and process dead time cannot be limited to a provisional level. It is that you do not get. However, if a good result can be obtained by the knowledge type adjustment method, a minimum amount of work is required.

  The knowledge type adjustment unit 20a of the control parameter adjustment unit 2a searches for the PID parameter by using the knowledge type adjustment method selected by the user in step S4-2 for the control target model generated by the simple modeling unit 11, and adjusts the adjustment information. The presentation unit 3a presents the PID parameter searched by the knowledge type adjustment unit 20a to the user (step S4-3 in FIG. 4). Since the operation of the knowledge type adjustment unit 20a is the same as that of the knowledge type adjustment unit 20 of the first embodiment, a detailed description thereof is omitted.

  FIG. 7 shows an example of an adjustment information presentation screen on which the adjustment information presentation unit 3a presents information in step S4-3. On the adjustment information presentation screen 700 shown in FIG. 7, an adjustment result (PID parameter adjustment value) 701 by the knowledge type adjustment unit 20a is presented. In addition, the adjustment information presentation screen 700 presents a description 702 of the modeling method of the simple modeling unit 11 as model information, and further presents information 703 on the transfer function expression of the control target model generated by the simple modeling unit 11. Yes.

  The adjustment information presentation screen 700 also includes a control response obtained as a result of executing a simulation for controlling the control target model generated by the simple modeling unit 11 by the PID controller that has adjusted the knowledge type PID parameter by the knowledge type adjustment unit 20a. An adjustment result graph 704 representing a knowledge type adjustment prediction response and a model response reproducing an actual control operation before adjustment by the knowledge type adjustment method is presented. The calculation of the knowledge type adjustment prediction response and the model response is performed by the adjustment information presentation unit 3a.

  Furthermore, the adjustment information presentation screen 700 has an adjustment completion selection button 705 for selecting completion of adjustment and an analysis type adjustment method when a good control result is not obtained by the knowledge type adjustment method. The analysis type adjustment technique selection button 706 is displayed.

  The user looks at the adjustment result graph 704 presented on the adjustment information presentation screen 700, confirms the control response predicted by the PID parameter by the knowledge type adjustment method, and sets the PID parameter as the adjustment result in the actual controller. Then, the control operation is executed. If it is determined that a good control result is obtained by this control operation, the adjustment completion selection button 705 is selected to complete the adjustment (NO in step S4-4 in FIG. 4). If the user determines that a good control result has not been obtained by the knowledge type adjustment method, the user selects analysis type adjustment method selection button 706 and proceeds to adjustment of analysis type (YES in step S4-4). .

  When the analysis type adjustment method is selected, the adjustment procedure control unit 4 causes the control operation base modeling unit 12 to perform modeling. The control action base modeling unit 12 is based on the modeling method disclosed in Japanese Patent Application Laid-Open No. 2004-38428. When the operation amount MV and the control amount PV are obtained from the control result of the actual control object based on the knowledge type adjustment. Series data is collected, modeling based on a realistic control operation is performed, and a control target model is generated (step S4-5 in FIG. 4). The modeling by the control action-based modeling unit 12 is based on the control action in which the characteristics of the control target are grasped even at a provisional level and the standard and safe knowledge type adjustment is performed. The probability that valid data will be collected increases.

  The analysis type adjustment unit 21a of the control parameter adjustment unit 2a searches for a PID parameter by using the analysis type adjustment method disclosed in Japanese Patent Application Laid-Open No. 2005-070940 for the control target model generated by the control action base modeling unit 12. Then, the adjustment information presentation unit 3a presents the PID parameters searched by the analysis type adjustment unit 21a to the user (step S4-6 in FIG. 4). Since the operation of the analysis type adjustment unit 21a is the same as that of the analysis type adjustment unit 21 of the first embodiment, detailed description thereof is omitted.

  FIG. 8 shows an example of an adjustment information presentation screen on which the adjustment information presentation unit 3a presents information in step S4-6. In the adjustment information presentation screen 800 shown in FIG. 8, an adjustment result (PID parameter adjustment value) 801 by the analysis type adjustment unit 21a is presented. In addition, on the adjustment information presentation screen 800, a description 802 of the modeling method of the control action base modeling unit 12 is presented as model information, and further, information 803 on the transfer function formula of the control target model generated by the control action base modeling unit 12 Is presented.

  Further, the adjustment information presentation screen 800 is obtained as a result of executing a simulation for controlling the control target model generated by the control action base modeling unit 12 by the PID controller in which the analysis type PID parameter is adjusted by the analysis type adjustment unit 21a. An adjustment result graph 804 representing an analysis-type adjustment prediction response that is a control response and a model response that reproduces an actual control operation before adjustment by the analysis-type adjustment method is presented. The calculation of the analytical adjustment prediction response and the model response is performed by the adjustment information presentation unit 3a.

  Further, on the adjustment information presentation screen 800, an adjustment completion selection button 805 for selecting adjustment completion and an empirical adjustment method are selected when a good control result is not obtained by the analysis type adjustment method. A rule-of-thumb adjustment method selection button 806 is displayed.

  The user looks at the adjustment result graph 804 presented on the adjustment information presentation screen 800, confirms the control response predicted by the PID parameter by the analysis type adjustment method, and sets the PID parameter as the adjustment result in the actual controller. Then, the control operation is executed. If it is determined that a good control result is obtained by this control operation, the adjustment completion selection button 805 is selected to complete the adjustment (NO in step S4-7 in FIG. 4). If the user determines that a good control result has not been obtained by the analytical adjustment method, he selects the empirical rule type adjustment method selection button 806 and proceeds to the empirical rule type adjustment (in step S4-7). YES).

  When the rule-of-thumb adjustment method is selected, the adjustment procedure control unit 4 causes the control task information input unit 13 to collect control tasks (trouble patterns). The user inputs a control task (for example, occurrence of hunting, delay of response, etc.) from the control result of the actual control object based on the analysis type adjustment. The control task information input unit 13 passes the control task input by the user to the empirical rule type adjustment unit 22 of the control parameter adjustment unit 2a.

  The rule-of-thumb adjustment unit 22 calculates a PID parameter readjustment value based on a PID parameter readjustment rule set in advance according to a control task (trouble pattern) selected by the user, and an adjustment information presentation unit 3a presents the PID parameter calculated by the rule-of-thumb adjustment unit 22 to the user (step S4-8 in FIG. 4). The empirical rule type readjustment performed by the empirical rule type adjusting unit 22 is disclosed in Japanese Patent Application Laid-Open No. 2000-148204, and a detailed description thereof will be omitted.

  The reason why such a rule-of-thumb re-adjustment method is necessary is that, because complete modeling is impossible due to the inevitable nonlinearity of the control target, etc., the final adjustment of the control error corresponding to the analysis error is required. is there. The range of control states that can be supported by heuristic-type readjustment is limited because of heuristics. The PID parameters must be adjusted in advance by an analytical adjustment method to the extent that one of the typical trouble patterns prepared in advance is applied.

  As described above, in this embodiment, every time processing is performed by any one of the knowledge type adjustment method, the analysis type adjustment method, and the heuristic type readjustment method, the PID parameter obtained by this processing is changed. Since the recommended value is presented, it is possible to cause the user to determine whether or not to continue searching for the recommended value of the PID parameter, and it is possible to further reduce the probability that the user will get stuck in the PID parameter adjustment.

[Third Embodiment]
Next, a third embodiment of the present invention will be described. This embodiment corresponds to Principle 2 of the invention described above. In the second embodiment, a user continuously registers a control target that is a determination criterion for ending the adjustment work, and continues determination that the adjustment work is continued when the actual control result does not satisfy the control target. An example in which steps are added will be described as a third embodiment. Here, the control target is, for example, a control result that can be calculated from time-series data (hereinafter referred to as control data) of the control amount PV and the operation amount MV during the control operation, such as an overshoot amount, rise time, and hunting amplitude. The feature value of

  In order to simplify the explanation, in this embodiment, the PID parameter adjustment support apparatus is provided with a logging function unit that automatically collects control data, and an item in which a user sets a control target from the control data collected by the logging function unit The feature value corresponding to is acquired, and the feature value is compared with the set value of the control target set by the user, so that the adjustment work is continued. Thereby, in this Embodiment, a user's input operation | work can be reduced compared with 2nd Embodiment.

  The logging function unit is for simplifying the explanation, and control data used for continuation determination of adjustment only needs to be input to the PID parameter adjustment support device. Of course, it does not matter if the user inputs. In other words, providing the logging function unit in the PID parameter adjustment support device is not an essential component requirement. In any case, the user input work related to the continuation determination of the adjustment is reduced.

FIG. 9 is a flowchart showing the flow of processing of the PID parameter adjustment support method according to the third embodiment of the present invention, and FIG. 10 shows the configuration of the PID parameter adjustment support apparatus according to the third embodiment of the present invention. It is a block diagram.
The PID parameter adjustment support apparatus according to the present embodiment includes an input information processing unit 1a, a control parameter adjustment unit 2a, an adjustment information presentation unit 3a, an adjustment procedure control unit 4, a knowledge type adjustment unit, and an analysis type adjustment described later. And an adjustment continuation determination unit 5 for determining whether or not to continue the search by the rule-of-thumb adjustment unit.

  The input information processing unit 1 a includes a simple modeling unit 11, a control action base modeling unit 12, and a control task information input unit 13. The control parameter adjustment unit 2a includes a knowledge type adjustment unit 20a, an analysis type adjustment unit 21a, and an empirical rule type adjustment unit 22.

  Hereinafter, the operation of the PID parameter adjustment support apparatus according to the present embodiment will be described with reference to FIG. First, the adjustment continuation determination unit 5 requests the user to set a control target. FIG. 11 shows an example of a control target setting screen 1100 in which the adjustment continuation determination unit 5 requests the user to set a control target. For example, the user sets 1 ° C. as the target value of the overshoot amount and 180 seconds as the target value of the rise time. Here, the item presented by the adjustment continuation determination unit 5 as a control target may be a feature value item that can be calculated from the control data, and is not limited to the overshoot amount or the rise time. The user only needs to set the target value for at least one type of item. For example, in the example of FIG. 11, only the rise time may be set.

  The processes in steps S10-2, S10-3, and S10-4 in FIG. 9 are the same as steps S4-1, S4-2, and S4-3 in FIG. However, in the present embodiment, since the adjustment continuation determining unit 5 determines whether or not to proceed to the analysis type adjustment, the adjustment information presenting unit 3a includes the adjustment information presenting screen 700 described in the second embodiment. The adjustment completion selection button 705 and the analysis type adjustment method selection button 706 are not displayed.

  The user sets a PID parameter, which is a knowledge type adjustment result, in the PID controller, and performs control of the control target. The adjustment continuation determination unit 5 receives control data (operation amount MV and control amount PV) during the control operation collected by the logging function unit (not shown) (step S10-5 in FIG. 9), and from this control data, the user Obtains a feature value corresponding to the item for which the control target is set (hereinafter, measured feature value), compares this measured feature value with the set value of the control target set by the user, and satisfies the control target. Is determined (step S10-6 in FIG. 9).

  The adjustment continuation determination unit 5 determines that the control target is satisfied if the measured feature value falls within a predetermined range based on the control target (YES in step S10-6 in FIG. 9), and the measured feature. The value and the control target are presented and the adjustment is finished (step S10-13 in FIG. 9). For example, if the measured feature value of the overshoot amount acquired from the control data is within the control target 1 ° C., the control target is satisfied, and if the measured feature value of the rise time is 180 seconds or less, the control target Is satisfied. In this case, for example, if the measured feature value of the overshoot amount is 0.5 ° C. and the measured feature value of the rise time is 175 seconds, it is determined that the control target is satisfied, and the process proceeds to step S10-13. become.

  On the other hand, the adjustment continuation determination unit 5 determines that the control target is not satisfied when the measured feature value is not within the predetermined range based on the control target (NO in step S10-6 in FIG. 9). The adjustment procedure control unit 4 is instructed to perform analysis type adjustment.

  The processes in steps S10-7 and S10-8 in FIG. 9 are the same as steps S4-5 and S4-6 in FIG. However, in the present embodiment, since the adjustment continuation determining unit 5 determines whether or not to proceed to the rule of thumb type adjustment, the adjustment information presenting unit 3a displays the adjustment information presenting screen 800 described in the second embodiment. Of these, the adjustment completion selection button 805 and the rule-of-thumb adjustment method selection button 806 are not displayed. Note that the control data collected in step S10-5 in FIG. 9 is used as the input data for modeling in step S10-7 in FIG.

  The user sets a PID parameter, which is an analysis-type adjustment result, in the PID controller, and controls the control target. The adjustment continuation determining unit 5 receives the control data during the control operation collected by the logging function unit (step S10-9 in FIG. 9), acquires the measured feature value from the control data, and the user can obtain the measured feature value. It is compared with the set value of the set control target to determine whether or not the control target is satisfied (step S10-10 in FIG. 9).

  The adjustment continuation determination unit 5 determines that the control target is satisfied if the actually measured feature value is within a predetermined range based on the control target (YES in step S10-10 in FIG. 9), and the actually measured feature. The value and the control target are presented and the adjustment is finished (step S10-13 in FIG. 9). The adjustment continuation determination unit 5 determines that the control target is not satisfied when the measured feature value is not within a predetermined range based on the control target (NO in step S10-10 in FIG. 9). The adjustment procedure control unit 4 is instructed to perform heuristic adjustment.

The process of step S10-11 in FIG. 9 is the same as step S4-8 in FIG.
The user sets a PID parameter, which is a rule-of-thumb adjustment result, in the PID controller, and performs control of the controlled object. The adjustment continuation determination unit 5 receives the control data during the control operation collected by the logging function unit (step S10-12 in FIG. 9), acquires the measured feature value from the control data, and acquires the measured feature value and the control target. And finishes the adjustment (step S10-13 in FIG. 9).

In addition, in the presentation of the control target achievement status in step S10-13 in FIG. 9, in addition to the presentation of the actual measurement feature value and the control target value, the determination result of whether or not the control target has been achieved, the adjustment type performed immediately before, It is preferable to present information regarding the adjustment method at the same time.
As described above, in this embodiment, since it is automatically determined whether or not the adjustment work is continued, the user's input work can be reduced as compared with the second embodiment. .

  The PID parameter adjustment support apparatuses in the first to third embodiments can be realized by, for example, a computer including a CPU, a memory, and an interface, and a program that controls these hardware resources. The CPU executes the processes described in the first to third embodiments according to the program stored in the memory.

  The present invention can be applied to a technology that supports adjustment of PID parameters of a PID controller.

  DESCRIPTION OF SYMBOLS 1, 1a ... Input information processing part, 2, 2a ... Control parameter adjustment part, 3, 3a ... Adjustment information presentation part, 4 ... Adjustment procedure control part, 5 ... Adjustment continuation judgment part, 10 ... Modeling part, 11 ... Simple modeling , 12 ... Control action base modeling part, 13 ... Control task information input part, 20, 20a ... Knowledge type adjustment part, 21, 21a ... Analysis type adjustment part, 22 ... Empirical rule type adjustment part.

Claims (9)

Input information processing means for acquiring information necessary for searching for PID parameters;
Knowledge type adjustment means for executing a knowledge type adjustment method to search for a recommended value of the PID parameter;
An analysis type adjustment means for executing an analysis type adjustment method to search for a recommended value of the PID parameter;
A rule-of-thumb adjustment means for executing a rule-of-thumb re-adjustment method to search for a recommended value of the PID parameter;
Adjustment procedure control means for causing at least one of the knowledge type adjustment means, the analysis type adjustment means, and the heuristic adjustment type means to execute processing;
An information presenting means for presenting a recommended value of the PID parameter obtained by this processing each time the processing by any of the knowledge type adjusting means, the analysis type adjusting means, and the heuristic type adjusting means is performed ,
The input information processing means includes
Simple modeling means for generating a control target model used by the knowledge type adjusting means using a simple modeling method;
A control action based modeling means for generating a control target model used by the analysis type adjusting means from a control result based on the adjustment of the knowledge type adjusting means;
Control task information input means for inputting a control task used by the rule-of-thumb adjustment means,
The adjustment procedure control means causes PID parameter adjustment to cause at least one means to execute processing in the order of the knowledge type adjustment means, the analysis type adjustment means, and the heuristic type adjustment means in accordance with a user instruction. Support device. Input information processing means for acquiring information necessary for searching for PID parameters;
Knowledge type adjustment means for executing a knowledge type adjustment method to search for a recommended value of the PID parameter;
An analysis type adjustment means for executing an analysis type adjustment method to search for a recommended value of the PID parameter;
A rule-of-thumb adjustment means for executing a rule-of-thumb re-adjustment method to search for a recommended value of the PID parameter;
Adjustment procedure control means for causing at least one of the knowledge type adjustment means, the analysis type adjustment means, and the heuristic adjustment type means to execute processing;
Information presenting means for presenting a recommended value of the PID parameter obtained by this processing each time the processing by any of the knowledge type adjusting means, the analysis type adjusting means, and the heuristic type adjusting means is performed;
The knowledge type adjustment means, the analysis type adjustment means, and a continuation determination means for determining whether or not to continue the search by the heuristic type adjustment means ,
The input information processing means includes
Simple modeling means for generating a control target model used by the knowledge type adjusting means using a simple modeling method;
A control action based modeling means for generating a control target model used by the analysis type adjusting means from a control result based on the adjustment of the knowledge type adjusting means;
Control task information input means for receiving a control task used by the rule-of-thumb adjustment means from a user,
The continuation determination unit compares the feature value obtained from the result of controlling the control target by the PID controller in which the recommended value of the PID parameter is set with the control target set by the user, thereby recommending the recommended value of the PID parameter. Whether or not the search of the knowledge type adjustment unit and the analysis type adjustment unit is completed,
The adjustment procedure control unit causes the at least one unit to execute processing in the order of the knowledge type adjustment unit, the analysis type adjustment unit, and the heuristic type adjustment unit according to the determination of the continuation determination unit. PID parameter adjustment support device. In the PID parameter adjustment support device according to claim 1 or 2 ,
The information presenting means further executes a simulation for controlling a controlled object by a PID controller in which a recommended value of the PID parameter is set, and presents a control response obtained as a result of the simulation. Support device. In the PID parameter adjustment support device according to any one of claims 1 to 3 ,
The PID parameter adjustment support apparatus, wherein the information presentation means further presents an explanation of each adjustment method type. In the PID parameter adjustment support device according to any one of claims 1 to 4 ,
The knowledge type adjustment method searches for a recommended value of the PID parameter by any one of the Ziegler-Nichols method, the CHR method, the IMC method, and the Cohen Coon method.
The analytical adjustment method calculates an evaluation function value indicating the closeness between a control response obtained by simulation and a predetermined ideal control response characteristic, and repeatedly executes the simulation while sequentially changing PID parameters. Then, the recommended value of the PID parameter for which the evaluation function value is the optimum value is searched for,
The heuristic-type readjustment method allows a user to select a control task generated in a control response obtained by simulation, and the PID parameter based on a readjustment rule set in advance according to the selected control task A PID parameter adjustment support apparatus characterized by searching for a recommended value. An input information processing step for acquiring information necessary for searching for PID parameters;
Knowledge type adjustment step for searching the recommended value of the PID parameter by executing the knowledge type adjustment method, Analysis type adjustment step for searching the recommended value of the PID parameter by executing the analysis type adjustment method, Re-adjustment of the heuristic type An adjustment procedure control step for performing at least one of heuristic adjustment steps for executing a technique and searching for a recommended value of the PID parameter;
An information presentation step for presenting a recommended value of a PID parameter obtained by this process each time the process according to any of the knowledge type adjustment step, the analysis type adjustment step, and the heuristic type adjustment step is performed ,
The input information processing step includes a simple modeling step of generating a control target model used by the knowledge type adjustment step using a simple modeling technique, and a control target model used by the analysis type adjustment step. Executing at least one of a control action base modeling step generated from a control result based on the adjustment of the control, a control task information input step of inputting a control task used by the heuristic type adjustment step,
In the adjustment procedure control step, at least one step is executed in the order of the knowledge type adjustment step, the analysis type adjustment step, and the heuristic type adjustment step in accordance with a user instruction. Support method. An input information processing step for acquiring information necessary for searching for PID parameters;
Knowledge type adjustment step for searching the recommended value of the PID parameter by executing the knowledge type adjustment method, Analysis type adjustment step for searching the recommended value of the PID parameter by executing the analysis type adjustment method, Re-adjustment of the heuristic type An adjustment procedure control step for performing at least one of heuristic adjustment steps for executing a technique and searching for a recommended value of the PID parameter;
An information presentation step for presenting a recommended value of the PID parameter obtained by this process every time the process according to any of the knowledge type adjustment step, the analysis type adjustment step, and the heuristic adjustment type step is performed;
The knowledge type adjustment step, the analysis type adjustment step, and a continuation determination step for determining whether or not to continue the search by the heuristic type adjustment step ,
The input information processing step includes a simple modeling step of generating a control target model used by the knowledge type adjustment step using a simple modeling technique, and a control target model used by the analysis type adjustment step. Executing at least one of a control action base modeling step generated from a control result based on the adjustment of the control, a control task information input step of inputting a control task used by the heuristic type adjustment step,
In the continuation determination step, the recommended value of the PID parameter is obtained by comparing the characteristic value obtained from the result of controlling the control target by the PID controller in which the recommended value of the PID parameter is set with the control target set by the user. Whether or not to continue the search every time the knowledge type adjustment step and the analysis type adjustment step are finished,
The adjustment procedure control step causes at least one step to execute processing in the order of the knowledge type adjustment step, the analysis type adjustment step, and the heuristic type adjustment step according to a determination result of the continuation determination step. PID parameter adjustment support method. The PID parameter adjustment support method according to claim 6 or 7 ,
The information presenting step further includes executing a simulation for controlling a controlled object by a PID controller in which a recommended value of the PID parameter is set, and presenting a control response obtained as a result of the simulation. Support method. The PID parameter adjustment support method according to any one of claims 6 to 8 ,
The information presentation step further presents an explanation of the type of each adjustment technique, and a PID parameter adjustment support method.

Images