Control method, control device and computer readable storage medium

Abstract

The present application discloses a control method, a control device and a computer-readable storage medium, and relates to the field of intelligent control. The present application obtains the historical correction matrix when the actuator executes the last instruction, and the target state difference after executing the last instruction; determines the target correction matrix based on the historical correction matrix; corrects the target state difference according to the target correction matrix; generates a target instruction based on the corrected target state difference, and the actuator is configured to execute the target instruction. The present application performs auxiliary correction on the state difference, such as the target state difference, through a correction matrix, such as a target correction matrix, and the correction matrix, such as the target correction matrix, approaches the inverse matrix of the actuator transfer matrix, thereby making the state difference, such as the target state difference, converge without divergence, and also improves the convergence speed of the state difference, such as the target state difference, and also expands the convergence space to the global, so that the stability is enhanced.

Description

Technical Field

[0001] The present application relates to the field of intelligent control, and in particular to a control method, a control device and a computer-readable storage medium. Background Art

[0002] In the case of changing the target from the current state to the target state under certain conditions, there are generally actuators and sensors that can judge the difference between the current state and the specified state. The sensor needs to wait until the actuator completes the action to measure the state, and use the sensor's perception information as feedback to control the actuator. Finally, the actuator reaches the target state after multiple executions. However, it may also cause execution divergence and deviate from the target state. Summary of the invention

[0003] In order to solve the above technical problems, a technical solution adopted in this application is: a control method, comprising:

[0004] Obtaining the historical correction matrix when the executor executed the last instruction, and the target state difference after executing the last instruction;

[0005] Determine a target correction matrix based on the historical correction matrix, wherein the target correction matrix is ​​closer to the inverse matrix of the actuator transfer matrix than the historical correction matrix;

[0006] Correcting the target state difference according to the target correction matrix;

[0007] A target instruction is generated based on the corrected target state difference amount, and the actuator is configured to execute the target instruction to be in a target state or a state close to the target state.

[0008] In order to solve the above technical problems, a technical solution adopted in this application is: a control device, comprising:

[0009] An acquisition module, used to acquire the historical correction matrix when the actuator executes the last instruction, and the target state difference after executing the last instruction;

[0010] A determination module, configured to determine a target correction matrix based on the historical correction matrix, wherein the target correction matrix is ​​closer to an inverse matrix of the actuator transfer matrix than the historical correction matrix;

[0011] a correction module, configured to correct the target state difference according to the target correction matrix; and

[0012] A generating module is used to generate a target instruction based on the corrected target state difference, and the executor is configured to execute the target instruction to achieve a state in the target state or a state close to the target state.

[0013] In order to solve the above technical problems, a technical solution adopted in the present application is: a computer-readable storage medium, in which program data is stored, and when the program data is executed by a processor, it is used to implement the method described above.

[0014] By adopting the above technical solution, the present application performs auxiliary correction on state difference quantities such as target state difference quantities through correction matrices such as target correction matrices. Correction matrices such as target correction matrices approach the inverse matrix of the actuator transfer matrix, thereby making state difference quantities such as target state difference quantities converge instead of diverging. In addition, the convergence speed of state difference quantities such as target state difference quantities is improved, and the convergence space is expanded to the global level, thereby enhancing stability. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] The drawings herein are incorporated into the specification and constitute a part of the specification. These drawings illustrate embodiments consistent with the present application and are used together with the specification to illustrate the technical solution of the present application.

[0016] Figure 1 This is a schematic diagram of the framework of the control device in some embodiments of the present application;

[0017] Figure 2 This is a flow chart of the control method in some embodiments of the present application;

[0018] Figure 3 Schematic diagram of the control method in other embodiments of the present application;

[0019] Figure 4 This is a flowchart of an iterative model in some embodiments of the present application;

[0020] Figure 5 This is a schematic diagram of the framework of the control device in some embodiments of the present application;

[0021] Figure 6 A schematic diagram of a computer-readable storage medium in some embodiments of the present application;

[0022] Figure 7 This is a schematic diagram of an electronic device in some embodiments of the present application. DETAILED DESCRIPTION

[0023] The present application is further described in detail below in conjunction with the accompanying drawings and examples. It is particularly noted that the following examples are only used to illustrate the present application, but are not intended to limit the scope of the present application. Similarly, the following examples are only some embodiments of the present application rather than all embodiments, and all other embodiments obtained by ordinary technicians in the field without making creative work are within the scope of protection of this application.

[0024] Reference to "embodiments" herein means that a particular feature, structure, or characteristic described in conjunction with the embodiments may be included in at least one embodiment of the present application. The appearance of the phrase in various locations in the specification does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment that is mutually exclusive with other embodiments. It is explicitly and implicitly understood by those skilled in the art that the embodiments described herein may be combined with other embodiments.

[0025] Next, a control device is described, which can change a target object from a current state to a target state. The control device can make the state difference converge without divergence, while increasing the convergence speed and expanding the convergence space to the whole world, thereby enhancing the system stability of the entire control device.

[0026] See also Figure 1 , Figure 1 Schematic diagram of the framework of the control device 100 in some embodiments of the present application. The control device 100 may include an actuator 10 for changing the state of a target object 11, a controller 20 for communicating with the actuator 10 to issue instructions to the actuator 10, and a sensor 30 for communicating with the controller 20 to feed back perception information to the controller 20. The controller 20 may generate instructions according to the perception information, and the actuator 10 may execute the instructions to change the state of the target object 11.

[0027] The perception information may include information about changes in the state of the actuator 10 and / or the target object 11 that the sensor 30 perceives during or after the actuator 10 executes the instruction. The information obtained by the sensor 30 sensing the actuator 10 and the information obtained by the sensor 30 sensing the target object 11 have correspondence, correlation, and mapping. Thus, when the information obtained by the sensor 30 sensing the actuator 10 is known, the information obtained by the sensor 30 sensing the target object 11 can be known. Similarly, when the information obtained by the sensor 30 sensing the target object 11 is known, the information obtained by the sensor 30 sensing the actuator 10 can be known.

[0028] The actuator 10 may have different structures or be adjustable according to different use environments and different objects. For example, the actuator 10 may be an electric actuator, a hydraulic actuator, a pneumatic actuator, or an electromagnetic actuator, etc., which can move the target object 11. In some embodiments, the actuator 10 may be a motor.

[0029] The controller 20 may be a single chip microcomputer, a computer, a mobile phone, or a server, and other processing devices that can perform logic operations, generate instructions, and process information. The specific selection can be made by those skilled in the art, and is not limited to any specific selection. The processing device may also be, for example, a game console, a personal computer, a tablet computer, a smart phone, or other types of processing devices.

[0030] There are many types of sensors 30, which can be selected or adjusted according to different usage environments. In some embodiments, the sensor 30 can be an ultrasonic sensor, an acceleration sensor, a magnetic sensor, an image sensor, a power sensor, or a displacement sensor, etc., and can be selected according to the needs of those skilled in the art without specific limitation. In some embodiments, the sensor 30 can be an image sensor such as a camera, a computer, or a mobile phone, etc., which has a camera function.

[0031] During the use of the control device 100, when the current state of the actuator 10 given by the sensor 30 is consistent with the target state, or when the current state of the target object 11 given by the sensor 30 is consistent with the target state, it can be considered that the actuator 10 or the target object 11 is in the target state and the control process is completed.

[0032] In order to describe the control device 100 more precisely, the following definitions of quantities are given:

[0033] 1. The behavior of controlling the actuator 10 is periodic, and the number of cycles that have passed at the current time is called the round number t (t is a positive integer);

[0034] 2. In the tth round, the target state difference u between the current state and the target state measured by the sensor 30 t (u t is an N-dimensional vector, N is a positive integer);

[0035] 3. In the tth round, the actuator obtains the instruction Δt (N-dimensional vector), which can change the current state accordingly after execution. For a linear system, the current state will increase by HΔt, where the transfer matrix H (N×N-dimensional matrix) is an unknown quantity that changes slowly or remains unchanged on this periodic time scale.

[0036] The most commonly used control process in the prior art is to approximate that the transfer matrix H is close to the unit matrix (unit matrix), corresponding to the situation where the actuator 10 and the sensor 30 are extremely ideal, so that the actuator 10 executes the instruction exactly with the target state difference u t Instead, we try to make the next round of state reach the target state. That is:

[0037] Δt=-u t

[0038] However, the actual situation is often not consistent with the ideal. The results after actual implementation are:

[0039] u t+1 =u t +HΔt=(IH)u t

[0040] Where I is the unit matrix. When the actuator 10 deviates from the ideal and there is an error, the convergence speed is a linear convergence speed, and there is still room for improvement. In addition, the control method used in the prior art also has problems in stability. When a certain eigenvalue of the transfer matrix H is less than 0 or greater than 2, divergence will occur, and then control failure will occur.

[0041] In this application, the initial correction matrix can be recorded as B0, and the correction matrix in the tth cycle can be recorded as B t .

[0042] For the tth round cycle, at the beginning, the sensor 30 first measures the target state difference u between the current state and the target state t , the change of the execution instruction of the actuator 10 is:

[0043] Δt=-B t u t

[0044] So the target state difference u at the beginning of the t+1 cycle t+1 for:

[0045] u t+1 =u t +HB t Δt=(I-HB t ) t

[0046] In order to minimize the target state difference, the final correction matrix B t To approach H -1 (the inverse matrix of the transfer matrix), then the update of the correction matrix also requires B t+1 Replace H as much as possible -1 , to minimize ||u t+2 ||. Among them, B t+1 B t To be closer to H -1 .

[0047] The present application improves the control method by modifying the state difference, such as the target state difference u, through a correction matrix, such as a target correction matrix. t Auxiliary correction is performed, and the correction matrix, such as the target correction matrix, approaches the inverse matrix of the transfer matrix of the actuator 10, thereby making the state difference, such as the target state difference u t Converges without divergence, and also improves the state difference, such as the target state difference u t The convergence speed is improved, and the convergence space is expanded to the global level, which enhances the stability.

[0048] In the present application, a correction matrix such as a target correction matrix can correct the influence of the transfer matrix H, and the convergence speed is superlinear. In addition, the convergence domain does not require that the eigenvalues ​​of the transfer matrix H must be within a certain range. As long as the transfer matrix H is reversible (that is, the control freedom of the actuator 10 is sufficient to satisfy the control freedom), it is globally converged.

[0049] Furthermore, a trivially differentiable system can be defined as: if there exists a parameter region Z in an n-dimensional parameter system with n-dimensional physical quantities, so that the Jacobi matrix under the parameters in any region Z is invertible, then the system is called a trivially differentiable system in region Z.

[0050] When the system is a trivially differentiable system, its local behavior can be approximated as a linear system with the Jacobian matrix as the transformation matrix. Furthermore, in this application, for some smooth and differentiable nonlinear systems (such as control problems involving the rotation of the actuator 10 or the target object 11), it can be equivalent to the transfer matrix H, which is the Jacobian matrix of the system, which changes with u t The control method of the present application is adaptable.

[0051] In some embodiments, the control device 100 may be used in a control process of photolithography alignment.

[0052] Next, a control method is described, which can be used for the control device 100 in the above embodiment. Figure 2 , Figure 2 Schematic diagram of the control method in some embodiments of the present application. In some embodiments, the controller 20 may implement the above method. In some embodiments, the controller 20 and the sensor 30 cooperate to implement the above method. The method may include:

[0053] Step S21: Obtain the historical correction matrix when the executor executes the last instruction.

[0054] Each time the actuator 10 executes an instruction, the correction matrix is ​​used to perform auxiliary correction on the state difference, and the correction matrix can be recorded and stored as a historical correction matrix.

[0055] Step S22: Obtain the target state difference after executing the last instruction.

[0056] The target state difference can be obtained through the sensing information of the sensor 30 .

[0057] In some embodiments, the perception information may be directly used as the target state difference. In some embodiments, step S22 may be implemented by the sensor 30. In some embodiments, the perception information obtained by the sensor 30 may be stored in the control device 100, and the controller 20 may implement step S22.

[0058] In some embodiments, the target state difference can be obtained by processing the perception information by the controller 20. In some embodiments, the controller 20 implements step S22.

[0059] In some embodiments, step S21 and step S22 may be performed in any order. In some embodiments, step S21 may be performed first, and then step S22. In some embodiments, step S22 may be performed first, and then step S21.

[0060] Step S23: Determine the target correction matrix based on the historical correction matrix.

[0061] Correction matrix B t Approaching H -1 , in order to replace H as much as possible -1 , to minimize ‖u t+1 ‖. Then the update of the correction matrix also requires B t+1 Replace H as much as possible -1 , to minimize ||u t+2 ||. Then the historical correction matrix can be used to determine. Then in order to make the state difference, such as the target state difference u t Converge without divergence, and require the target correction matrix to be closer to the inverse matrix of the actuator's transfer matrix than the historical correction matrix. Based on this restriction, the state difference can also be improved, such as the target state difference u t The convergence speed is improved, and the convergence space is expanded to the global level, which enhances the stability.

[0062] In some embodiments, the historical correction matrix may be iteratively processed based on the iterative model to obtain the target correction matrix. Specifically, the historical correction matrix may be input into the iterative model, and then the iterative model outputs the target correction matrix.

[0063] In some embodiments, step S23 is implemented by the controller 20 .

[0064] Step S24: Correct the target state difference according to the target correction matrix.

[0065] The target state difference u at the beginning of the t+1 cycle t+1 for:

[0066] u t+1 =u t +HB t Δt=(I-HB t ) t

[0067] By correcting, we can minimize ||u t+1 ||.

[0068] In some embodiments, step S24 is implemented by the controller 20 .

[0069] Step S25: Generate a target instruction based on the corrected target state difference amount.

[0070] The target instruction is executed by the actuator 10, and the state is brought into the target state or close to the target state.

[0071] In some embodiments, step S25 is implemented by the controller 20 .

[0072] In some embodiments, after step S25, the target state has not been reached, and then after step S25, step S21 and step S22 may continue to be executed (step S21 and step S22 may be performed in any order), and then the cycle is continued until the target state difference meets the convergence criteria (i.e., is in the target state).

[0073] See also Figure 3 , Figure 3 Schematic diagram of the control method in some other embodiments of the present application. The method may also include:

[0074] Step S31: Use the non-zero matrix as the history correction matrix.

[0075] Before the executor 10 executes the instruction for the first time, there may be no correction matrix, and then a correction matrix may be specified or preset. In some embodiments, the specified or preset correction matrix may be a non-zero matrix, that is, a non-zero matrix is ​​used as a historical correction matrix. In some embodiments, the specified or preset correction matrix may be a unit matrix.

[0076] In some embodiments, step S31 is implemented by the controller 20 .

[0077] Step S32: Obtain the target state difference of the actuator.

[0078] Before the executor 10 executes the instruction for the first time, there may not be the execution of the previous instruction, and the target state difference can be directly obtained. Specifically, step S32 can refer to step S22 and will not be described in detail. In some embodiments, the target state difference in step S32 can be obtained by using the acquisition method of step S22. In some embodiments, step S31 and step S32 can be performed in any order. In some embodiments, step S31 can be performed first, and then step S32. In some embodiments, step S32 can be performed first, and then step S31.

[0079] In some embodiments, step S32 is implemented by the controller 20 .

[0080] In some embodiments, step S31 and step S32 may be performed first (step S31 and step S32 may be performed in any order), and then step S23, step 24, and step 25 may be performed in sequence, and then step S21, step S22 (step S21 and step S22 may be performed in any order), step S23, step S24, and step S25 may be performed in a loop until the target state difference meets the convergence criteria.

[0081] See also Figure 4 , Figure 4 The following is a flow chart of an iterative model in some embodiments of the present application. The method for establishing the iterative model may include:

[0082] Step S41: Obtain the target state difference after executing the last instruction.

[0083] Specifically, step S41 can refer to step S22, and will not be described in detail. In some embodiments, step S41 can be replaced by step S22. In some embodiments, step S22 can be replaced by step S41.

[0084] In some embodiments, step S41 is implemented by the controller 20 .

[0085] Step S42: Obtain the historical state difference when the executor executes the last instruction.

[0086] Specifically, step S42 can refer to step S22, and will not be described in detail. In some embodiments, the historical state difference amount in step S42 can be obtained by using the acquisition method of step S22.

[0087] In some embodiments, step S42 is implemented by the controller 20 .

[0088] Step S43: In response to the target instruction being the result of the first instruction, a non-zero matrix that is not equal to the target state difference amount is used as the historical state difference amount.

[0089] Before the actuator 10 executes the instruction for the first time, there may be no historical state difference amount, and then a historical state difference amount may be specified or preset. In some embodiments, the specified or preset historical state difference amount may be a non-zero matrix, that is, a non-zero matrix is ​​used as the historical state difference amount. In some embodiments, the specified or preset historical state difference amount may be a unit matrix.

[0090] In some embodiments, step S43 is implemented by the controller 20 .

[0091] Step S44: Obtain the target state difference of the actuator.

[0092] Before the executor 10 executes the instruction for the first time, there may not be the execution of the previous instruction, and the target state difference amount can be directly obtained. Specifically, step S44 can refer to step S22, which is not described in detail. In some embodiments, the target state difference amount in step S44 can be obtained by using the acquisition method of step S22. In some embodiments, step S44 can refer to step S32, which is not described in detail. In some embodiments, step S44 can be replaced by step S32. In some embodiments, step S32 can be replaced by step S44.

[0093] In some embodiments, step S44 is implemented by the controller 20 .

[0094] Step S45: Determine an iteration model based on the historical state difference amount and the target state difference amount.

[0095] The historical state difference and the target state difference are used as parameters of the iterative model, and then the correction matrix can be updated by updating the historical state difference and the target state difference, so that the correction matrix B t Approaching H -1 , in order to replace H as much as possible -1 , to minimize u t+1 ‖. Then the update of the correction matrix also requires B t+1 Replace H as much as possible -1 , to minimize ‖u t+2 ‖. Then the state difference, such as the target state difference u t Converge without divergence, and require the target correction matrix to be closer to the inverse matrix of the actuator's transfer matrix than the historical correction matrix. Based on this restriction, the state difference can also be improved, such as the target state difference u t The convergence speed is improved, and the convergence space is expanded to the global level, which enhances the stability.

[0096] In some embodiments, step S45 is implemented by the controller 20 .

[0097] In some embodiments, step S41 and step S42 may be performed in any order and cooperate with each other to facilitate the smooth performance of step S45. In some embodiments, step S41 may be performed first and then step S42. In some embodiments, step S42 may be performed first and then step S41.

[0098] In some embodiments, step S43 and step S44 may be performed in any order and cooperate with each other to facilitate the smooth performance of step S45. In some embodiments, step S43 may be performed first and then step S44. In some embodiments, step S44 may be performed first and then step S43.

[0099] In some embodiments, the iteration model is:

[0100]

[0101] (u t+1 -u t ) T u t ≠0

[0102] Where t+1 is a positive integer; B t is the t-th correction matrix, t is a positive integer; u t is the state difference at the tth time, and I is the unit matrix.

[0103] Step S45 may include: taking the historical state difference as u t , the target state difference is taken as u t+1 Generate iterative models;

[0104] Then step S23 may include: taking the historical correction matrix as B t Input the iterative model and output B t+1 As the target correction matrix.

[0105] In some embodiments, based on the target state difference u at the beginning of the t+1 round cycle t+1 for:

[0106] u t+1 =u t +HB t Δt=(I-HB t ) t

[0107] Therefore, in order to minimize the state difference, the final correction matrix B t To approach H -1 , then the update of the correction matrix also requires B t+1 Replace H as much as possible -1 , to minimize ‖u t+1 ‖. Since the only information that can be relied on recently is u obtained in round t+1 t+1 , so that B t+1 The following conditions must be met:

[0108]

[0109] It is worth pointing out that such an N-dimensional vector equation is not sufficient to determine the N×N-dimensional matrix B t+1 , that is, B t+1In some embodiments, the Moore-Penrose pseudoinverse method can be used to directly obtain B t+1 , but it is necessary to find the inverse after the singular value decomposition, and each step requires a large amount of calculation. t+1 There is an order of magnitude reduction during the convergence process, and the calculation process often involves subtracting two close numbers, which can easily cause machine errors.

[0110] In some embodiments, a single rank transform is defined as follows:

[0111] The set of all m×n-dimensional matrices in the number field K is Mat{K:m×n}, mapping If satisfied:

[0112]

[0113] (rankP = 1 and Q(M) = M(I+P))

[0114] The mapping Q is called an n-dimensional single-rank transformation on the number field K.

[0115] When rankP=1, P can be described by a dyadic vector, and its inverse algorithm can use the Sherman-Morrison formula (the Sherman-Morrison formula is named after Jack Sherman and Winifred J. Morrison. In linear algebra, it is a method for solving the inverse matrix) to avoid the machine error problem caused by subtracting similar numbers.

[0116] Furthermore, we can order:

[0117]

[0118] After simplification, it becomes:

[0119]

[0120] Using the Sherman-Morrison formula, for any column vector u, v, we have:

[0121]

[0122] The iterative model can be obtained:

[0123]

[0124] Since the actual execution is in a normal convergence situation, there is almost no situation where two approximate numbers are subtracted in the above iterative model, thereby eliminating the influence of machine errors.

[0125] In some embodiments, when the denominator in the iterative model is 0, there are three possibilities:

[0126] 1.u t =0, then the target state has been reached, control is completed, and no further operation is required. In some embodiments, the method further includes: in response to the result that the target state difference amount is 0, generating end information indicating that control is completed (or in the target state).

[0127] 2.u t+1 -u t =0, then the actuator 10, such as the motor, is not moving, and the degree of freedom is insufficient, and the actuator 10, such as the motor, needs to be repaired in hardware. In some embodiments, the method further includes: in response to the result that the target state difference minus the historical state difference is 0, generating a warning message of an actuator operation failure.

[0128] 3.(u t+1 -u t )⊥u t , then at this time, the execution of the actuator 10, such as the motor, and the instruction are not in the same degree of freedom, and the degree of freedom needs to be exchanged. In some embodiments, the method also includes: generating an error message of an instruction error in response to the result perpendicular to the historical state difference amount after subtracting the historical state difference amount from the target state difference amount.

[0129] In some embodiments, the pseudo code is:

[0130]

[0131] Since this application only requires simple inner products and matrix multiplications in each cycle, and does not involve inverse operations, and the only additional storage space required is the correction matrix corresponding to the size of the degrees of freedom, the calculation is very simple and fast, and can be used in conjunction with embedded systems.

[0132] In this application, the following beneficial effects are achieved:

[0133] Convergence speed: Due to B t As the iteration proceeds, it gradually approaches H -1 , so ||I-HB t+1 ||<<||Iu t+1 ||, the convergence rate is superlinear.

[0134] Stability: Due to B t As the iteration proceeds, it gradually approaches H -1 , so (I-HB t ) approaches 0, thus globally converging and being independent of the eigenvalue of H.

[0135] In some embodiments, the method can be used in a control process of lithography alignment.

[0136] Next, a control device is described. The control device can be used to implement the above method and can also be used in the control device 100 in the above embodiment. In some embodiments, the control device 200 can be the control device 100 in the above embodiment. In some embodiments, the control device 200 can be the controller 20 in the above embodiment. Figure 5 , Figure 5 Schematic diagram of the framework of the control device in some embodiments of the present application. The control device 200 may include an acquisition module 201, a determination module 202, a correction module 203 and a generation module 204. In some embodiments, the acquisition module 201 may not be a part of the control device 100. In some embodiments, the control device 200 may also include the actuator 10 in the above embodiment.

[0137] In some embodiments, the acquisition module 201 can be used to acquire the historical correction matrix when the actuator executes the last instruction, and the target state difference after executing the last instruction. In some embodiments, the acquisition module 201 can be used to execute steps S21, S22 and S42 in the above embodiments. In some embodiments, the acquisition module 201 can be the sensor 30 in the above embodiments.

[0138] In some embodiments, the determination module 202 may be used to determine a target correction matrix based on the historical correction matrix. The target correction matrix is ​​closer to the inverse matrix of the actuator's transfer matrix than the historical correction matrix. In some embodiments, the determination module 202 may be used to perform step S23 in the above embodiment.

[0139] In some embodiments, the correction module 203 may be used to correct the target state difference according to the target correction matrix. In some embodiments, the correction module 203 may be used to execute step S24 in the above embodiment.

[0140] In some embodiments, the generation module 204 may be used to generate a target instruction based on the corrected target state difference. The actuator 10 in the above embodiment may execute the target instruction to perform a state in the target state or a state close to the target state. In some embodiments, the generation module 204 may be used to perform step S25 in the above embodiment.

[0141] In some embodiments, the acquisition module 201 may be used to use the non-zero matrix as the history correction matrix. In some embodiments, the acquisition module 201 may be used to perform step S31 in the above embodiment.

[0142] In some embodiments, the acquisition module 201 may be used to acquire the target state difference of the actuator. In some embodiments, the acquisition module 201 may be used to execute step S32 and step S44 in the above embodiments.

[0143] In some embodiments, the determination module 202 may be configured to iteratively process the historical correction matrix based on an iterative model to obtain a target correction matrix.

[0144] In some embodiments, the acquisition module 201 may be used to acquire the historical state difference when the executor executes the last instruction. In some embodiments, the acquisition module 201 may be used to execute step S41 in the above embodiment.

[0145] In some embodiments, the acquisition module 201 can be used to respond to the target instruction being the result of the first instruction and use a non-zero matrix that is not equal to the target state difference as the historical state difference. In some embodiments, the acquisition module 201 can be used to execute step S43 in the above embodiment.

[0146] In some embodiments, the control device 200 may further include an iteration model generation module 205. The iteration model generation module 205 may be used to determine the iteration model based on the historical state difference amount and the target state difference amount. In some embodiments, the iteration model generation module 205 may be a part of the determination module 202.

[0147] In some embodiments, the iteration model is:

[0148]

[0149] (u t+1 -u t ) T u t ≠0

[0150] Where t+1 is a positive integer; B t is the t-th correction matrix, t is a positive integer; u t is the state difference at the tth time, and I is the unit matrix.

[0151] In some embodiments, the iterative model generation module 205 may be used to take the historical state difference as u t , the target state difference is taken as u t+1 Generate iterative models;

[0152] In some embodiments, the determination module 202 may be configured to use the historical correction matrix as B t Input the iterative model and output B t+1 As the target correction matrix.

[0153] In some embodiments, the determination module 202 may be configured to determine a result of subtracting a historical state difference amount from a target state difference amount.

[0154] In some embodiments, the generating module 204 may be configured to generate a warning message of an actuator operation failure in response to the determination module 202 outputting a result that the target state difference minus the historical state difference is 0.

[0155] In some embodiments, the determination module 202 may be configured to determine whether the target state difference is zero.

[0156] In some embodiments, the generating module 204 may be configured to generate target state completion completion information in response to the determining module 202 outputting a result that the target state difference amount is 0.

[0157] In some embodiments, the determination module 202 may be used to determine whether the target state difference amount minus the historical state difference amount is perpendicular to the historical state difference amount;

[0158] In some embodiments, the generation module 204 may be configured to generate error information of an instruction error in response to a result of the determination module 202 outputting the target state difference amount minus the historical state difference amount, which is perpendicular to the historical state difference amount.

[0159] Next, a computer-readable storage medium is described. Figure 6 , Figure 6 Schematic diagram of a computer-readable storage medium in some embodiments of the present application. The computer-readable storage medium 300 is used to store program data 301. When the program data 301 is executed by the processor, it is used to implement the method in any of the above embodiments. The computer-readable storage medium 300 can also be used in the above-mentioned control device 100 and control device 200. In some embodiments, the computer-readable storage medium 300 can be the memory in the above-mentioned embodiments.

[0160] The computer-readable storage medium 300 may specifically be a medium that can store program data 301, such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, or may be a server that stores the program data 301. The server may send the stored program data 301 to other devices for execution, or may execute the stored program data 301 by itself.

[0161] Next, an electronic device is described. Figure 7 , Figure 7Schematic diagram of an electronic device in some embodiments of the present application. The electronic device can be applied to the above method. The electronic device 300 includes a processor 401 and a memory 402 coupled to each other. The memory 402 stores program data, which is used to implement the method in any of the above embodiments when executed by the processor 401. The memory 402 can be the memory in the above embodiment.

[0162] In some embodiments, the processor 401 is used to run the program data stored in the memory 402 to implement the method of any of the above embodiments. Specifically, the processor 401 controls the operation of the electronic device, and the processor 401 can also be called a CPU (Central Processing Unit). The processor 401 may be an integrated circuit chip with signal and image processing capabilities. The processor 401 can also be a general-purpose processor, a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA) or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components. The general-purpose processor can be a microprocessor or the processor can also be any conventional processor, etc.

[0163] In some embodiments, the memory 402 is used to store program data executed by the processor 401 and data during the processing of the processor 401, wherein the memory 402 may include a non-volatile storage portion for storing the above program data. In another embodiment, the memory 402 may only be used as the memory of the processor 401 to cache the data during the processing of the processor 401, and the program data is actually stored in a device outside the processor 401. The processor 401 connects to the external device and calls the externally stored program data to perform the corresponding processing.

[0164] In some embodiments, the program data may be the program data 301 in the above embodiments.

[0165] In some embodiments, the memory 402 may be Figure 6 The computer readable storage medium 300 in.

[0166] In the several embodiments provided in this application, it should be understood that the disclosed methods and devices can be implemented in other ways. For example, the device implementation described above is only illustrative, for example, the division of modules or units is only a logical function division, and there may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed.

[0167] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the present embodiment.

[0168] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0169] The above description is only an implementation method of the present application, and does not limit the patent scope of the present application. Any equivalent structure or equivalent process transformation made using the contents of the present application specification and drawings, or directly or indirectly used in other related technical fields, are also included in the patent protection scope of the present application.

Claims

1. A control method, characterized in that: include: Obtaining the historical correction matrix when the executor executed the last instruction, and the target state difference after executing the last instruction; Determining a target correction matrix based on the historical correction matrix, wherein the target correction matrix is ​​closer to the inverse matrix of the actuator transfer matrix than the historical correction matrix, wherein determining the target correction matrix based on the historical correction matrix comprises: iteratively processing the historical correction matrix based on an iterative model to obtain the target correction matrix; Correcting the target state difference according to the target correction matrix; generating a target instruction based on the corrected target state difference, the actuator being configured to execute the target instruction to be in a target state or a state approaching the target state; Wherein, the iterative model is: (in t+1 -in t ) T in t ≠0 Wherein, t+1 is a positive integer; B t is the t-th correction matrix, t is a positive integer; u t is the state difference at the tth time, I is the unit matrix; The method further comprises: Obtaining a historical state difference amount when the executor executes the last instruction, or, in response to the target instruction being the result of the first instruction, taking a non-zero matrix that is not equal to the target state difference amount as the historical state difference amount; The historical state difference is taken as u t , the target state difference is taken as u t+1 generating the iterative model; The iterative processing of the historical correction matrix based on the iterative model to obtain the target correction matrix includes: The history correction matrix is ​​taken as B t Input the iterative model and output B t+1 as the target correction matrix.

2. The method according to claim 1, characterized in that When the target instruction is a first instruction, before determining the target correction matrix based on the historical correction matrix, the method further includes: The non-zero matrix is ​​used as the history correction matrix, and the target state difference of the actuator is obtained.

3. The method according to claim 1, characterized in that The method further comprises: Obtaining a historical state difference amount when the executor executes the last instruction, or, in response to the target instruction being the result of the first instruction, taking a non-zero matrix that is not equal to the target state difference amount as the historical state difference amount; The iteration model is determined based on the historical state difference amount and the target state difference amount.

4. The method according to claim 1 or 2, characterized in that: The method further comprises: Obtaining a historical state difference when the executor executes the last instruction; In response to a result that the target state difference amount minus the historical state difference amount is 0, warning information of the actuator operation failure is generated.

5. The method according to claim 1 or 2, characterized in that: The method further comprises: In response to the result that the target state difference amount is 0, end information in the target state is generated.

6. The method according to claim 1 or 2, characterized in that: The method further comprises: Obtaining a historical state difference amount when the executor executes the last instruction, or, in response to the target instruction being the result of the first instruction, taking a non-zero matrix that is not equal to the target state difference amount as the historical state difference amount; In response to a result perpendicular to the historical state difference amount after the historical state difference amount is subtracted from the target state difference amount, error information of an instruction error is generated.

7. A control device, characterized in that: include: An acquisition module, used to acquire the historical correction matrix when the actuator executes the last instruction, and the target state difference after executing the last instruction; A determination module, configured to determine a target correction matrix based on the historical correction matrix, wherein the target correction matrix is ​​closer to an inverse matrix of the actuator transfer matrix than the historical correction matrix; A correction module, used for correcting the target state difference according to the target correction matrix; as well as a generating module, configured to generate a target instruction based on the corrected target state difference, wherein the executor is configured to execute the target instruction to be in a target state or a state close to the target state; The determination module is used to iteratively process the historical correction matrix based on an iterative model to obtain the target correction matrix; The iterative model is: (in t+1 -in t ) T in t ≠0 Wherein, t+1 is a positive integer; B t is the t-th correction matrix, t is a positive integer; u t is the state difference at the tth time, I is the unit matrix; The acquisition module is used to acquire the historical state difference amount when the executor executes the last instruction, or, in response to the target instruction being the result of the first instruction, to use a non-zero matrix that is not equal to the target state difference amount as the historical state difference amount; The control device may further include an iterative model generation module, wherein the iterative model generation module is used to use the historical state difference as u t , the target state difference is taken as u t+1 generating the iterative model; The determining module is used to use the historical correction matrix as B t Input the iterative model and output B t+1 as the target correction matrix.

8. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores program data, and when the program data is executed by a processor, it is used to implement the method according to any one of claims 1 to 6.

Images