Feedback control device
By using the moment of inertia adjustment and database-driven control method in the feedback control device, the safety and comfort of the feedback control method in the prior art on equipment such as hydraulic excavators are solved, and the comfort and safe operation of the equipment are achieved. .
Patent Information
- Application Number
- CN202080082965.3
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Priority Date
- 2019-12-05
- Filing Date
- 2020-11-18
- Publication Date
- 2024-12-06
- Estimated Expiration
- 2040-11-18
AI Technical Summary
The existing feeling feedback control method requires frequent adjustment of the PID gain to cope with the nonlinearity of people's feeling when handling the operation of hydraulic excavators and other equipment, and the equipment may automatically operate when no one is operating, resulting in safety hazards.
By adjusting the moment of inertia as a parameter, the database-driven control method is used to determine the target angular velocity and control input based on the operation input and comfort difference, ensuring the comfort and safety of the operator.
It realizes comfortable and safe operation on hydraulic excavators and other equipment, avoids automatic actions when unmanned operations, and improves the operator's comfort and equipment control accuracy.
Smart Images

Figure CN114730162B_ABST
Abstract
Description
Technical Field
[0001] The invention relates to a feedback control device. Background Art
[0002] According to a survey conducted by the Cabinet Office, Japan has a high GDP but a low happiness level. In other words, there is a huge gap between the "material abundance" associated with GDP and the "physical and mental abundance" associated with happiness.
[0003] As a strategy to fill this gap, it has been proposed to improve the richness of the body and mind by allowing already highly developed "material" (cars, welfare support equipment, etc.) to act in a way that takes human sensitivity into consideration.
[0004] Most of the research on perception is about static fields such as product design evaluation and design, but the inventors of the present invention proposed a "perception feedback control method" for dynamically controlling perception (for example, Non-Patent Document 1).
[0005] The conventional sensory feedback control method is composed of a cascade control system, which generates a target speed of the controlled device using a PID controller by providing the desired comfort. At this time, in order to cope with the nonlinearity of human sensory perception, an effective database-driven control method is used for nonlinear systems (for example, non-patent documents 2 and 3).
[0006] However, the conventional "feeling feedback control method" requires the PID gain to be adjusted every time the target speed is changed in the human brain, so the target speed may change in various ways in the operating system of a hydraulic excavator, etc. In addition, in the structure of the control system, the hydraulic excavator, etc., may automatically operate even when no human is operating, so another countermeasure is required to prevent this from happening.
[0007] Prior art literature
[0008] Non-patent literature
[0009] Non-patent literature 1: T.Kinoshita and T.Yamamoto, "Design of a Data-OrientedKansei Feedback Control System", Journal of Robotics, Networking and ArtificialLife, Vol.4-No.1,14 / 17(2017)
[0010] Non-patent document 2: T. Yamamoto, K. Takao and T. Yamada, "Design of a Data-Driven PID Controller", IEEE Transon control systems Technology, Vol. 17-No. 1, 29 / 39 (2009)
[0011] Non-patent document 3: Shin Wakitani, Yoshihiro Onishi, and Toru Yamamoto, “Design of nonlinear PID control system using FRIT method”, Measurement and Control, Vol. 52-No. 10, 885 / 891 (2013). Summary of the invention
[0012] Here, an object of the present invention is to provide a feedback control device that allows an operator to operate a device comfortably and safely.
[0013] To achieve the above-mentioned purpose, the inventors of the present invention have studied intensively and have come up with the present invention, that is, for a device that is at least rotationally operated by an operator, the moment of inertia of the transfer function is focused on and the moment of inertia is adjusted as a parameter to improve the comfort of the operator.
[0014] A feedback control device according to one aspect of the present invention includes: an object device that is at least rotated by an operation input of an operator; a first target angular velocity determination unit that determines a target angular velocity of an output angular velocity of the object device based on the operation input using a first transfer function that is a transfer function of the object device, the first transfer function being defined by a first moment of inertia representing the moment of inertia of the object device; a control unit that determines a control input to the object device based on a difference between the first target angular velocity and the output angular velocity; a second target angular velocity determination unit that determines a second target angular velocity that is a target angular velocity corresponding to the operation input requested by the operator based on the operation input; a comfort level determination unit that determines a comfort level of the operator based on a difference between the output angular velocity and the second target angular velocity; and a database that sequentially accumulates the first target angular velocity, the comfort level, and the target comfort level, wherein the first target angular velocity determination unit adjusts the first moment of inertia in such a manner that the difference between the target comfort level and the comfort level becomes smaller using the first target angular velocity, the comfort level, and the target comfort level accumulated in the database. BRIEF DESCRIPTION OF THE DRAWINGS
[0015] Figure 1 It is a configuration diagram of a feedback control device according to an embodiment.
[0016] Figure 2 Schematic diagram showing the relationship between the comfort level and the output error of the feedback control device according to the embodiment.
[0017] Figure 3 This is a schematic diagram showing the results of simulating the time changes of the comfort level, the output angular velocity, and the operation input in the feedback control using the feedback control device according to the embodiment.
[0018] Figure 4 This is a schematic diagram showing how the moment of inertia is adjusted in feedback control using the feedback control device according to the embodiment. DETAILED DESCRIPTION
[0019] Hereinafter, a feedback control device according to an embodiment of the present invention will be described with reference to the accompanying drawings. In addition, the technical protection scope of the present invention is not limited to the following embodiments, and any changes can be made within the scope of the technical concept of the present invention.
[0020] Figure 1 FIG. 1 is a diagram showing a structure of a feedback control device 100 according to the first embodiment. Figure 1 As shown, the feedback control device 100 includes a target device 10, a target angular velocity determination unit 20 (first target angular velocity determination unit), a control unit 30, a comfort level determination unit 40, and a database 50. The feedback control device 100 also includes an intra-brain target angular velocity determination unit 70 (second target angular velocity determination unit), an operating mechanism 60, subtractors 81, 82, 83, and a target comfort level setting unit 90.
[0021] The target device 10 is inputted by the operator through operation. h The target device 10 is a rotating drive unit of, for example, an automobile or a construction machine. The rotating drive unit is, for example, a boom of a hydraulic excavator. The target device 10 operates at an output angular velocity yω corresponding to a control input τ input from the control unit 30. The target device 10 inputs the output angular velocity yω to the subtractor 82 and feeds the output angular velocity yω back to the subtractor 81.
[0022] The target angular velocity determination unit 20 determines the target angular velocity based on the operator's operation input τ to the target device 10. h The target angular velocity rω (first target angular velocity) of the output angular velocity yω of the target device 10 is determined using the first transfer function which is the transfer function of the target device 10 . The first transfer function has a first moment of inertia representing the moment of inertia of the target device 10 .
[0023] Specifically, the target angular velocity determination unit 20 uses the target comfort level r, the comfort level y, and the target angular velocity rω stored in the database 50 to adjust the first moment of inertia as a parameter so as to reduce the difference between the target comfort level r and the comfort level y. The target angular velocity determination unit 20 inputs the operation input τ to the first transfer function in which the first moment of inertia is adjusted. h , determines the target angular velocity rω.
[0024] The operation mechanism 60 receives input of the operator's operation on the target device 10. Specifically, the operation mechanism 60 includes an operation lever for driving the rotation drive unit, and the operation from the operator is input to the operation lever. h The target angular velocity is input to the target angular velocity determination unit 20 and the target angular velocity determination unit 70 in the brain.
[0025] The subtractor 81 calculates a difference obtained by subtracting the output angular velocity yω from the target angular velocity rω, and inputs the difference to the control unit 30 .
[0026] The control unit 30 determines the control input τ to the target device 10 by, for example, PID control based on the difference between the target angular velocity rω and the output angular velocity yω. Specifically, the control unit 30 determines the control input τ by PID control so that the difference input from the subtractor 81 becomes small.
[0027] The subtractor 82 calculates the target angular velocity r h The difference e after subtracting the output angular velocity yω h (speed deviation), and the difference e h The result is input to the comfort level determination unit 40 .
[0028] The comfort level determination unit 40 is based on the difference e h The comfort level y is determined. The comfort level determination unit 40 inputs the comfort level y into the database 50 and the subtractor 83 and outputs the comfort level y to the outside.
[0029] The subtractor 83 calculates the difference between the target comfort level r and the comfort level y, and inputs the difference to the target angular velocity determination unit 20 .
[0030] The database 50 sequentially accumulates the target angular velocity rω, the comfort level y, and the target comfort level r of the comfort level y.
[0031] The target comfort level setting unit 90 sets a target comfort level r, and inputs the set target comfort level r into the database 50 and the subtractor 83. The target comfort level r uses, for example, a predetermined value.
[0032] In the feedback control device 100 of this embodiment, the functions of the target angular velocity determination unit 20, the control unit 30, the comfort level determination unit 40, the intracerebral target angular velocity determination unit 70, and the target comfort level setting unit 90 are implemented by causing a computer to execute a program. The computer includes a processor that operates according to the program and a memory that stores data required for executing the program as main hardware components.
[0033] Furthermore, in the feedback control device 100 of the present embodiment, the database 50 is constituted by a computer-readable storage medium such as a ROM, an optical disk, or a hard disk drive.
[0034] One of the features of the feedback control device 100 of the present embodiment is that the target angular velocity determination unit 20 determines the target angular velocity rω by adjusting the first moment of inertia as a parameter in such a manner that the difference between the target comfort level r and the comfort level y becomes smaller, using the target comfort level r, the comfort level y, and the target angular velocity rω accumulated in the database 50. The method for adjusting the first moment of inertia will be described in detail in the embodiment described later.
[0035] For example, assuming that the first moment of inertia is J and the viscosity coefficient is D, the first transfer function is (1 / D) / (1+(J / D)·s). In this case, 1 / D is the system gain. For example, the stable angular velocity during boom operation of a full-lever input hydraulic excavator is easy to measure. Here, the viscosity coefficient D can also be used as a known fixed value, and the target angular velocity determination unit 20 determines the target angular velocity rω. Alternatively, the target angular velocity determination unit 20 can also determine the target angular velocity rω while adjusting the viscosity coefficient D together with the first moment of inertia J as parameters.
[0036] In the present application, the first moment of inertia of the target device 10 is denoted as “J”, and the first moment of inertia adjusted as a parameter by the target angular velocity determination unit 20 is denoted as “J ref ”.
[0037] Furthermore, the comfort level determination unit 40 may also obtain the vital sign response of the operator, such as the heart rate, through a sensor, etc., and determine the comfort level y corresponding to the operator based on the vital sign response. Alternatively, the comfort level determination unit 40 may determine the comfort level y corresponding to the operator using the comfort level personally input by the operator using an input device such as a touch panel.
[0038] The target angular velocity determination unit 70 in the brain determines the target angular velocity based on the input operation τ h , determines the operation input τ requested by the operator h The corresponding target angular velocity is the target angular velocity r h (Second target angular velocity). Target angular velocity r hFor example, the operator's own operation input τ h The corresponding target angular velocity (desired angular velocity). h The difference between the output angular velocity yω and h is the main factor that determines the operator's comfort level y. For example, at the target angular velocity r h The difference between the output angular velocity yω and h When it is 0, the comfort level y is determined to have the highest value. h 0 can also indicate that the operator's operation of the actual object device 10 has reached a level that can obtain the same output angular velocity yω as the target angular velocity rh. Here, in the feedback control device 100 in actual application, the target angular velocity r h and the difference e h The value of is unknown. Moreover, the brain target angular velocity determination unit 70, by inputting the operation τ h Input to the second transfer function, the target angular velocity r can be calculated h .
[0039] For example, assuming that the second moment of inertia representing the moment of inertia of the target device 10 requested by the operator is J h , the viscosity coefficient is D, and the second transfer function is (1 / D) / (1+(J h Here, the second moment of inertia J h The value of is unknown.
[0040] The feedback control device 100 of the present embodiment described above includes the target angular velocity determination unit 20 and the control unit 30. The target angular velocity determination unit 20 determines the target angular velocity based on the operation input τ to the target device 10. h The control unit 30 determines the target angular velocity rω of the output angular velocity yω of the target device 10; the control unit 30 determines the control input τ to the target device 10 based on the difference between the target angular velocity rω and the output angular velocity yω. The feedback control device 100 further includes an intra-brain target angular velocity determination unit 70 and a comfort determination unit 40, wherein the intra-brain target angular velocity determination unit 70 determines the control input τ based on the input operation τ. h Determine the target angular velocity r requested by the operator h The comfort level determination unit 40 is based on the output angular velocity yω and the target angular velocity r h The difference between h The target angular velocity determination unit 20 determines the comfort level y of the operator. The target angular velocity determination unit 20 reduces the difference between the target comfort level r and the comfort level y while adjusting the first moment of inertia J ref The target angular velocity rω is determined while being adjusted as a parameter.
[0041] Here, the comfort level is based on the target angular velocity r h The difference between the output angular velocity yω and h Therefore, the comfort level determination unit 40 can determine the comfort level to have the highest value when the operator successfully performs an operation to make the target angular velocity of the target device requested by the operator the same as the actual target angular velocity of the target device. Therefore, the operator can operate the target device 10 comfortably.
[0042] Furthermore, according to the feedback control device 100 of the present embodiment, only when there is an operation input τ h The target angular velocity rω is determined, the control input τ to the target device 10 is determined, and the target device 10 is operated. Therefore, it is possible to avoid the situation where the target device 10 automatically starts to operate when the operator does not operate the target device 10, and safety can be ensured.
[0043] As described above, according to the present embodiment, it is possible to provide the feedback control device 100 that allows an operator to operate a device comfortably and safely.
[0044] In the feedback control device 100 of the present embodiment, it is assumed that the first moment of inertia is J ref , the viscosity coefficient is D, the first transfer function can be expressed as (1 / D) / (1+(J ref / D)·s).
[0045] Therefore, the feedback control device 100 can accurately feedback the rotational motion of the control target device 10. In this case, the target angular velocity determination unit 20 may determine the target angular velocity rω by setting the viscosity coefficient D as a fixed value.
[0046] Therefore, the parameter to be adjusted is only the first moment of inertia J ref , which can achieve one-parameter tuning. On the other hand, the target angular velocity determination unit 20 can also combine the viscosity coefficient D with the first moment of inertia J ref The target angular velocity rω is determined while being adjusted together as a parameter. In this case, the feedback control device 100 can more accurately feedback the rotational motion of the device 10 to be controlled by taking into account the temperature dependency of the viscosity coefficient D and the like.
[0047] Furthermore, in the feedback control device 100 of the present embodiment, the target machine 10 may be a boom of a hydraulic excavator.
[0048] In this case, the boom of the hydraulic excavator can be operated comfortably and safely.
[0049] In the feedback control device 100 of the present embodiment, the intra-brain target angular velocity determination unit 70 has a function of determining the target angular velocity r corresponding to the operation input τh. h The second transfer function. Assume that the second moment of inertia is J h , the viscosity coefficient is D, and the second transfer function is (1 / D) / (1+(J h / D)·s). The brain target angular velocity determination unit 70 uses the adjusted first moment of inertia J ref The value of the second moment of inertia J is set h .
[0050] In this case, the first moment of inertia J can be used ref To determine the unknown second moment of inertia J h .
[0051] Example
[0052] The following, Figure 1 1 and 2. The embodiment (simulation example) of the feedback control device 100 shown in FIG. 1 is described. In this embodiment, the target device 10 is the boom of a hydraulic excavator, and the operator's comfort level y for boom operation can be measured. A "target comfort level r" is given as a target value for the comfort level y, and the output angular velocity yω of the boom of the hydraulic excavator is controlled so that the comfort level y satisfies the "target comfort level r".
[0053] The purpose of this embodiment is to improve the comfort y(t) (t is time). It is not easy to directly set the appropriate target angular velocity rω(t) of the target device 10 (the boom of the hydraulic excavator) at each moment. Here, according to the operation input (the joystick input torque in this embodiment) τ h (t) and the first moment of inertia J ref The relationship between is shown in the following formula (1), which can automatically generate an appropriate target angular velocity rω(t) at each moment.
[0054]
[0055] In formula (1), J ref is the first moment of inertia of the closed-loop transfer function including the boom of the hydraulic excavator, and D is the viscous resistance. As shown in formula (1), 1 / D is the system gain. Since the stable angular velocity when the boom of the hydraulic excavator is operated with full-lever input can be easily measured, the viscosity coefficient D is treated as a known fixed value in this embodiment.
[0056] In this embodiment, the first moment of inertia J is adjusted based on the database driven method described later. ref , in order to improve the comfort level y. Here, it is assumed that the second moment of inertia requested in the operator's brain is Jh When the first moment of inertia J ref Equal to the second moment of inertia J h , the actual moment of inertia of the control system is consistent with the moment of inertia in the brain. Therefore, assuming that the first moment of inertia J ref The state equal to the second moment of inertia Jh is the most comfortable state. In other words, when the operator successfully performs the same operation as the target response speed of the hydraulic excavator requested by the operator in his mind, the comfort level y indicates the most comfortable value.
[0057] Furthermore, in this embodiment, the comfort level y(t) of the control object is a discrete-time nonlinear system represented by the following formula (2).
[0058] y(t)=f(φ(t-1))…(2)
[0059] In formula (2), y(t) is the system output. f() is a nonlinear function. φ(t-1) is an information vector representing the state of the system before time t. The information vector φ(t-1) is defined by the following formula (3).
[0060] φ(t-1):=[y(t-1),…,y(tn y ),
[0061] r ω (t-1),…,r ω (tn w )]…(3)
[0062] Here, rω(t) (target angular velocity of the target device 10) is a system input in the control of the comfort level y(t). y is the number of system outputs. n w is the number of system inputs. In database-driven control, each operation and operation data is accumulated in the form of formula (3) in the database 50 (refer to Figure 1 ). In addition, the information vector φ(t) representing the current state of the system is called a request point (query).
[0063] Next, the first moment of inertia J is controlled by the database driven control performed by the target angular velocity determination unit 20 using the data accumulated in the database 50. ref The following section describes the adjustments made by the
[0064] Initial database creation
[0065] In database-driven control, if there is no past accumulated data, the first moment of inertia J cannot be adjusted in principle. ref Therefore, it is necessary to use the initial first moment of inertia Jref Input / output data is acquired, and an initial database (a part of the database 50) consisting of an information vector φ expressed by the following formula (4) is created.
[0066]
[0067] In formula (4), j = 1, 2, ..., N (0), It is given by the following formula (5).
[0068]
[0069] Furthermore, N(0) represents the number of initial data (the number of information vectors in the initial database). ref is fixed, so J ref (1) = J ref (2) = ... = J ref (N(0)).
[0070] Distance calculation, neighbor selection
[0071] Demand Points and the information vector accumulated in the database The distance between them is calculated by the weighted L1 norm represented by the following formula (6).
[0072]
[0073] In formula (6), j = 1, 2, ..., N(t), N(t) represents the number of data (information vectors) accumulated in the database at time t. represents the first element of the jth information vector. Similarly, represents the first feature of the demand point at time t. And, Represents all the information vectors present in the database The greatest element of the first element, Indicates the smallest feature among the first features.
[0074] In this embodiment, k information vectors obtained by formula (6) are selected from the side with the smaller distance d, and the selected data set is defined as a neighbor.
[0075] Local controller configuration
[0076] Next, for the neighbors selected as described above, a local controller as shown in the following formula (7) is constructed by the linearly weighted average method (Linearly Weighted Average: LWA).
[0077]
[0078] Here, wi is the value of J contained in the selected i-th information vector. ref The weight of (i) is given by the following formula (8).
[0079]
[0080] Through the above steps, the first moment of inertia J at each time t can be calculated refold (t). In addition, in order to adjust the first moment of inertia J more appropriately through a database-driven method ref , it is necessary to learn the database 50. Here, in this embodiment, the J of each data set in the database 50 is performed by the steepest descent method. ref Updates.
[0081] The first moment of inertia J is calculated by the steepest descent method. ref Update
[0082] In this embodiment, the steepest descent method of formula (9) is used to calculate the first moment of inertia J ref adjustments.
[0083]
[0084] In formula (9), η is a learning coefficient, and I(t) represents an evaluation criterion defined by the following formulas (10) and (11).
[0085]
[0086] ε(t):=r(t)-y(t))…(11)
[0087] Furthermore, the partial differential of the second term on the right side of formula (9) can be expanded as described in formula (12).
[0088]
[0089] By repeatedly implementing the steepest descent method until the evaluation criteria defined by formula (10) and formula (11) become sufficiently small, the optimized first moment of inertia J can be calculated. ref , that is, the second moment of inertia J in the operator's brain can be approximated h The first moment of inertia J is calculated by ref .
[0090] In addition, in this embodiment, the human perception model is set based on Weber-Fechner's law (IP Herman: Physics of the Human Body: Biological and Medical Physics, Biomedical Engineering, Springer-Verlag GmbH & CO.KG (2007)). This law expands Weber's law and is derived by assuming that there is a logarithmic function relationship between the sensation amount F and the stimulus amount R, with k as a constant, as described in the following formula (13).
[0091] F=k log R…(13)
[0092] The Weber-Fechner law has the following characteristics, that is, as the intensity of the stimulus gradually increases, the intensity of the sensation changes dramatically for the initial stimulus, but gradually its rate of increase slows down, which corresponds to the intuitive fact (Eitaro Masuyama: "Sensory Examination from a Psychological Perspective", Sensory Examination and Psychology, Vol.22-No.3, 107 / 114 (1986)). For example, in an experiment of lifting a weight of 100g, the feeling will be very different if you start from a state where you are already holding a 1kg weight and start from a state where you are not holding anything. From this, it can be confirmed that the feeling becomes duller as the weight becomes heavier. The specific formula of the human perception model is explained in the following numerical example.
[0093] Numerical example
[0094] Hereinafter, a numerical example of the feedback control according to the above-mentioned embodiment will be described.
[0095] exist Figure 1 The feedback control device 100 shown is assumed to be a boom of a hydraulic excavator and is given by a first-order lag system represented by the following formula (14).
[0096]
[0097] On the other hand, the comfort level y(t) is expressed by the following formula (15) and formula (16) using the Weber-Fechner law so that the maximum value of the comfort level y(t) becomes 1.
[0098]
[0099] e h (t) = r h (t)-y w(t)…(16)
[0100] In formula (15) and formula (16), r h (t) is the target angular velocity of the boom of the hydraulic excavator that the operator has in mind, and is calculated by the following formula (17).
[0101]
[0102] In formula (17), J h is the second moment of inertia of the boom of the hydraulic excavator in the operator's brain, which is unknown in this numerical example. h (t) is the speed error (target angular velocity r) of the hydraulic excavator boom perceived by the operator. h and the difference between the output angular velocity yω).
[0103] If the difference e in the operator's brain h (t) is completely zero, and according to formula (15), the comfort level y(t) becomes the maximum 1. In addition, E(t) is a variable related to the comfort level y(t), and has different values depending on the operator. The E(t) corresponding to the operator can be determined based on the vital sign response of the vital sign sensor or the comfort level y input by the operator. Figure 2 is the difference e in the operator's brain h The relationship between (t) and comfort y(t). Figure 2 As shown, the difference e h The larger E(t) is, the lower the comfort level y(t) is. Also, the larger E(t) is, the greater the rate of decrease in comfort level y(t) is.
[0104] In addition, it is assumed that the setting parameters in the numerical example described below are r(t)=1.0 (fixed to the maximum value 1 of the comfort level y(t)), η=0.8, E(t)=1, and D=1.
[0105] Furthermore, in this numerical example, it is assumed that the boom of a hydraulic excavator is raised or lowered. At this time, due to gravity, the boom speed will be faster when the boom is lowered than when the boom is raised. Therefore, it is assumed that the operator also has such a premise in his mind. As shown in the following formula (18), it is assumed that the second moment of inertia J h (t) varies according to a predetermined operation input (joystick input torque) τh(t). Under this assumption, it is confirmed that the first moment of inertia J obtained by the adjustment method described in the embodiment ref (t) has a second moment of inertia J h (t) followability.
[0106]
[0107] In addition, regarding the input / output relationship of the control unit 30, for example, assuming that K P , K I , K D Using these as PID parameters, PID control is performed as shown in the following formula (19).
[0108]
[0109] ε(t) in formula (19) is the difference between the target angular velocity rω(t) and the output angular velocity yω(t), as shown in the following formula (20).
[0110] ε(t)=r ω (t)-y ω (t)…(20)
[0111] The PID parameters of the control unit 30 can use existing parameters corresponding to the target device 10. The existing parameters can be fixed constants or variable parameters. In this numerical example, it is assumed that K P =10, K I =0.5, K D =0.
[0112] Under the above settings, the first moment of inertia J is adjusted by database driven control. ref (t), while performing the comfort y(t), output angular velocity yω(t) and operation input (joystick input) τ h Simulation of the time variation of (t). Figure 3 is a schematic diagram showing the simulation results. Figure 3 The upper graph of is a graph showing the temporal change between the comfort level y(t) and the target comfort level r(t). In the upper graph, the solid line shows the comfort level y(t), and the dotted line shows the target comfort level r(t). Figure 3 The graph in the middle section shows the output angular velocity yω(t) and the target angular velocity r h In the middle graph, the solid line represents the output angular velocity yω(t), and the dotted line represents the target angular velocity rh(t). Figure 3 The graph in the lower part shows the operation input τ h A graph of the time changes of . Figure 4 is the first moment of inertia J ref Schematic diagram of the adjusted condition. Figure 4 In the figure, the solid line represents the first moment of inertia J ref , the dotted line represents the second moment of inertia J h .in addition, Figure 3 The r shown h (t) (the target angular velocity of the boom of the hydraulic excavator that the operator has in mind) is obtained by adding the above-mentioned set values and the second moment of inertia J shown in formula (18) h Substitute (t) into formula (17) to obtain the calculated result.
[0113] from Figure 3 It can be confirmed that the comfort level (system output) y(t) is following the target comfort level r(t) (=1). This is because the output angular velocity yω(t) is following the target angular velocity r h (t) can also be confirmed. At this time, the second moment of inertia J h (t) According to the operation input (joystick input) τ h (t) varies, but Figure 4 It can be seen that through the database driven control of the embodiment, the first moment of inertia J ref (t) Follow the second moment of inertia J h (t) method, and appropriately estimate the second moment of inertia J h (t). This shows that by applying the database-driven control of the embodiment, the operator's second moment of inertia J can be inferred. h .
[0114] According to the embodiments and numerical examples described above, it is confirmed that, in the case of the boom of a hydraulic excavator as the object equipment 10 operated by the operator, the operator's comfort level is based on the Weber-Fechner law, and the first moment of inertia J is considered to be ref The control system is implemented by adjusting the first moment of inertia J ref Single parameter tuning can improve the comfort of hydraulic excavator boom operation.
[0115] The above embodiments (including examples) of the present invention have been described, but the present invention is not limited to the above embodiments, and various changes can be made within the scope of the invention. That is, the description of the above embodiments is only an example in nature and is not intended to limit the present invention, the application of the present invention, or the use of the present invention.
[0116] For example, in the above-mentioned embodiment, the feedback control device 100 is described by taking the boom of a hydraulic excavator as an example, but the present invention is also applicable to other equipment that is at least rotated by an operator.
[0117] Furthermore, in the above embodiment, the viscosity coefficient D in the formula (1) is used as a fixed value and the first moment of inertia J is calculated. ref However, as an alternative, the first moment of inertia J can also be tuned refand parameter tuning of the viscosity coefficient D. Furthermore, the transfer function of the target device 10 is not limited to (1 / D) / (1+(J / D)·s). For example, if viscosity does not need to be considered, a transfer function that does not include the viscosity coefficient D can be used. Alternatively, if loss needs to be considered, a transfer function that includes a loss term can also be used.
[0118] Furthermore, the feedback control device 100 may also be a database-driven type control of the target angular velocity determination unit 20 that utilizes a learning function. In this case, as the operator's operation time on the object device 10 becomes longer, the object device 10 gradually becomes capable of performing an output response that is more suitable for the operator. As a result, the operator's comfort level y can be further improved. In particular, when the learning is further deepened and a database 50 customized for each operator can be created, the first moment of inertia J can be adjusted based on the contents of the database 50 and sensor input from the object device 10, etc., instead of using the comfort level y. ref .
[0119] Summary of implementation methods
[0120] A feedback control device according to one aspect of the present invention includes: an object device that is at least rotated by an operation input of an operator; a first target angular velocity determination unit that determines a target angular velocity of an output angular velocity of the object device based on the operation input using a first transfer function that is a transfer function of the object device, the first transfer function being defined by a first moment of inertia representing the moment of inertia of the object device; a control unit that determines a control input to the object device based on a difference between the first target angular velocity and the output angular velocity; a second target angular velocity determination unit that determines a second target angular velocity that is a target angular velocity corresponding to the operation input requested by the operator based on the operation input; a comfort level determination unit that determines a comfort level of the operator based on a difference between the output angular velocity and the second target angular velocity; and a database that sequentially accumulates the first target angular velocity, the comfort level, and the target comfort level, wherein the first target angular velocity determination unit adjusts the first moment of inertia in such a manner that the difference between the target comfort level and the comfort level becomes smaller using the first target angular velocity, the comfort level, and the target comfort level accumulated in the database.
[0121] According to this configuration, based on the operation input to the object device, the target angular velocity of the output angular velocity of the object device, i.e., the first target angular velocity, is determined. Based on the difference between the first target angular velocity and the output angular velocity, the control input to the object device is determined. Based on the difference between the second target angular velocity, i.e., the target angular velocity corresponding to the operation input requested by the operator, and the output angular velocity, the comfort of the operator is determined. The first moment of inertia is adjusted in such a way that the difference between the target comfort and the comfort becomes smaller. In this way, the first target angular velocity can be determined while adjusting the first moment of inertia of the first transfer function in such a way that the comfort of the operator is improved, and the control input can be determined based on the difference between the first target angular velocity and the output angular velocity.
[0122] Here, the comfort level is determined based on the difference between the second target angular velocity corresponding to the operation input requested by the operator and the output angular velocity. For this reason, when the operator successfully performs an operation that makes the target angular velocity of the object device requested by the operator the same as the actual target angular velocity of the object device, the comfort level determination unit can determine the comfort level to have the highest value. Therefore, the operator can comfortably operate the object device.
[0123] Furthermore, according to the feedback control device of this embodiment, only when there is an operation input, the first target angular velocity is determined, the control input to the object device is determined, and the object device is operated. Therefore, it is possible to avoid the situation where the object device automatically starts to operate when the operator does not operate the object device, and safety can be ensured.
[0124] In the feedback control device, the first transfer function may also be defined by a viscosity coefficient. Assuming that the first moment of inertia is J ref , the viscosity coefficient is D, the first transfer function is (1 / D) / (1+(J ref / D)·s).
[0125] According to this configuration, the rotational motion of the device to be controlled can be accurately feedback-controlled.
[0126] In the feedback control device, the first target angular velocity determination unit may adjust the first moment of inertia by setting the viscosity coefficient D as a fixed value.
[0127] According to this configuration, since the parameter to be adjusted is only the first moment of inertia, single-parameter tuning can be achieved.
[0128] In the feedback control device, the first target angular velocity determination unit may determine the first target angular velocity while further adjusting the viscosity coefficient D as a parameter.
[0129] According to this configuration, the rotational movement of the device to be controlled can be fed back more accurately.
[0130] In the feedback control device, the target device may be a boom of a hydraulic excavator.
[0131] According to this configuration, the boom of the hydraulic excavator can be operated comfortably and safely.
[0132] In the feedback control device, the second target angular velocity determination unit may have a second transfer function for determining the second target angular velocity corresponding to the operation input, the second transfer function being defined by a second moment of inertia representing the moment of inertia of the target device requested by the operator and the viscosity coefficient, assuming that the second moment of inertia is J h , the viscosity coefficient is D, and the second transfer function is (1 / D) / (1+(J h / D)·s), the second moment of inertia J h The first moment of inertia J is adjusted ref The value of .
[0133] According to this configuration, the unknown second moment of inertia can be determined using the first moment of inertia.
[0134] In the feedback control device, the comfort level determination unit may set the comfort level so that the smaller the difference between the output angular velocity and the second target angular velocity, the greater the comfort level.
[0135] According to this configuration, when the operator successfully performs an operation to make the target angular velocity of the target device requested by the operator the same as the actual target angular velocity of the target device, the comfort level can be set to a relatively high value.
Claims
1. A feedback control device, characterized in that include: The target device is at least rotationally operated by an operator's operation input; a first target angular velocity determination unit that determines, based on the operation input, a first target angular velocity that is a target angular velocity of an output angular velocity of the target device using a first transfer function that is a transfer function of the target device, the first transfer function being defined by a first moment of inertia that represents a moment of inertia of the target device; a control unit that determines a control input to the target device based on a difference between the first target angular velocity and the output angular velocity; a second target angular velocity determination unit that determines, based on the operation input, a second target angular velocity, which is the target angular velocity corresponding to the operation input requested by the operator; a comfort level determination unit that determines a comfort level of the operator based on a difference between the output angular velocity and the second target angular velocity; as well as, database, accumulating the first target angular velocity, the comfort level and the target comfort level in sequence, wherein: The first target angular velocity determination unit adjusts the first moment of inertia by using the first target angular velocity, the comfort level, and the target comfort level accumulated in the database so as to reduce a difference between the target comfort level and the comfort level.
2. The feedback control device according to claim 1, characterized in that: The first transfer function is also defined by the viscosity coefficient, When the first moment of inertia is J ref , when the viscosity coefficient is D, the first transfer function is (1 / D) / (1+(J ref / D)·s).
3. The feedback control device according to claim 2, characterized in that: The first target angular velocity determination unit adjusts the first moment of inertia by setting the viscosity coefficient D as a fixed value.
4. The feedback control device according to claim 2, characterized in that: The first target angular velocity determination unit determines the first target angular velocity while further adjusting the viscosity coefficient D as a parameter.
5. The feedback control device according to any one of claims 1 to 4, characterized in that: The target equipment is a boom of a hydraulic excavator.
6. The feedback control device according to claim 2, characterized in that: The second target angular velocity determination unit has a second transfer function for determining the second target angular velocity corresponding to the operation input. The second transfer function is defined by a second moment of inertia representing the moment of inertia of the target device requested by the operator and the viscosity coefficient. Assume that the second moment of inertia is J h , the viscosity coefficient is D, and the second transfer function is (1 / D) / (1+(J h / D)·s), The second moment of inertia J h The first moment of inertia J is adjusted ref The value of .
7. The feedback control device according to any one of claims 1 to 4, characterized in that: The comfort level determination unit sets the comfort level in such a manner that the smaller the difference between the output angular velocity and the second target angular velocity, the greater the comfort level.
Citation Information
Patent Citations
Inertia ratio adjusting method based on reference model self-adaption
CN104656556A
Environmental comfort level control system and method
CN105717960A
Electric power steering device
CN108778904A
Comfort-based self-driving vehicle speed control method
CN109311478A
Interior Comfort HVAC User-Feedback Control System And Apparatus
US20160223215A1
