Control method and system of suspension type gravity compensation device

Abstract

The invention relates to a control method and system of a suspension type gravity compensation device. The method comprises the following steps: acquiring tension and inclination angle data of a suspension rope at each moment; calculating an expected gravity compensation value based on the inclination angle of the suspension rope at the current moment, and calculating the deviation between the suspension force and the expected gravity compensation value at the current moment; taking the deviation and the change rate of the deviation relative to time as input variables of a fuzzy PID control algorithm, taking a correction value of a PID control parameter as an intermediate output variable of the fuzzy PID control algorithm, and determining a basic domain of discourse of each variable; according to the determined basic discourse domain and the input variables, calculating discourse domain scaling factors of each input variable and the intermediate output variable at corresponding moments, and dynamically adjusting the discourse domain range of the corresponding variable by using the calculated discourse domain scaling factors; and generating a PID control quantity based on the adjusted discourse domain range. The control precision and the response speed of the suspension type gravity compensation device can be effectively improved.

Description

Technical Field

[0001] The present invention relates to the technical field of space gravity unloading, and particularly to a control method and system for a suspended gravity compensation device. Background Art

[0002] For the ground gravity unloading test of a data transmission antenna pointing mechanism, currently, actuators such as motors, electric cylinders, and pneumatic springs are usually used to provide rope tension, and servo-type closed-loop tension control is adopted to achieve active suspension. In the horizontal direction, a passive following mechanism or an active position tracking mechanism needs to be designed to follow the suspension point position to simulate the two-dimensional space deployment of the pointing mechanism. If a passive following mechanism is adopted, there is an elastic link in the vertical direction and a long suspension rope in the horizontal direction for this device, which is a typical underactuated system. Therefore, it is often difficult to meet the test requirements in terms of gravity unloading effect and following accuracy. Although the fuzzy PID controller can achieve a basically feasible control effect, in order to further improve the control accuracy of the system, the fuzzy domain can be subdivided, that is, the number of fuzzy subsets is increased, but this method will lead to redundant fuzzy rules, low utilization rate, and reduce the real-time performance of the control system. At the same time, its fuzzy domain, quantization factor, and proportional factor are all constants and cannot be changed once set. When the actual gravity unloading working condition parameters change, the response speed will slow down and the adaptability will become worse during the control process. Summary of the Invention

[0003] The technical problem to be solved by the present invention is to provide a control method and system for a suspended gravity compensation device, which can effectively improve the control accuracy and response speed of the suspended gravity compensation device.

[0004] The technical solution adopted by the present invention to solve its technical problems is: to provide a control method for a suspended gravity compensation device, including the following steps:

[0005] Obtain the tension and inclination data of the suspension ropes at each moment;

[0006] Calculate the expected gravity compensation value based on the inclination of the suspension rope at the current moment, and calculate the deviation e between the suspension force and the expected gravity compensation value at the current moment;

[0007] Use the deviation e and the change rate ec of the deviation e with respect to time as the input variables of the fuzzy PID control algorithm, and use the correction value of the PID control parameter as the intermediate output variable of the fuzzy PID control algorithm, and determine the basic domains of each of the input variables and the intermediate output variable;

[0008] According to the determined basic domains and the input variables, calculate the domain scaling factors of each of the input variables and the intermediate output variable at the corresponding moment, and use the calculated domain scaling factors to dynamically adjust the domain ranges of the corresponding variables;

[0009] Generate the PID control quantity based on the adjusted universe of discourse range.

[0010] Furthermore, the universe of discourse scaling factor of the input variable is calculated by the following formula:

[0011]

[0012] where α E and α EC are the universe of discourse scaling factors of the input variable e and the input variable ec respectively, E and EC are the boundary values of the basic universes of discourse of the input variable e and the input variable ec respectively, τ E and τ EC are the scaling factor coefficients of the input variable e and the input variable ec respectively, and ε is a positive number approaching 0.

[0013] Furthermore, the scaling factor coefficient is selected based on the integral optimal criterion.

[0014] Furthermore, the universe of discourse scaling factor of the intermediate output variable is calculated by the following formula:

[0015]

[0016] where β KP , β KI , β KD are the universe of discourse scaling factors of the output variables ΔK P , the output variables ΔK I , the output variables ΔK D respectively, U KP , U KI , U KD are the boundary values of the basic universes of discourse of the output variables ΔK P , the output variables ΔK I , the output variables ΔK D respectively, a i , b i and c i are all tuning parameters, and i = 1, 2, 3.

[0017] Furthermore, the tuning parameters are obtained by fitting through the linear autoregressive method using multiple sets of known working condition parameters of the suspended gravity compensation device.

[0018] Furthermore, the expected gravity compensation value is calculated by the following formula:

[0019] F d0 = F d (1 + sinθ)

[0020] where Fd0 is the expected gravity compensation value, F d is the load gravity applied to the suspended gravity compensation device, and θ is the inclination angle of the suspension rope.

[0021] Furthermore, the suspension force is the vertical component of the rope tension.

[0022] Furthermore, generating the PID control quantity based on the adjusted universe range includes:

[0023] Based on the adjusted universe range, respectively construct the first rule base for the proportional link and the differential link, and the second rule base for the integral link according to the set quantization level and membership function, and perform fuzzy inference;

[0024] Use the centroid method to defuzzify the fuzzy set to obtain the defuzzified intermediate output variable, and then generate the PID control quantity using the corrected PID control parameters.

[0025] The present invention also provides a control system for a suspended gravity compensation device, including:

[0026] A force sensor for detecting the tension of the suspension rope;

[0027] An angle sensor for detecting the inclination angle of the suspension rope;

[0028] A controller, including a parameter setting module, a data acquisition module, a gravity compensation calculation module, and an expected force adjustment module, where:

[0029] The data acquisition module is used to collect the tension and inclination angle data;

[0030] The expected force adjustment module is used to calculate the expected gravity compensation value at the current moment according to the inclination angle data collected in real time and the gravity load applied to the suspended gravity compensation;

[0031] The gravity compensation calculation module is used to execute any of the above methods;

[0032] The motion control module is used to drive the expansion and contraction of the electric cylinder according to the output control quantity of the compensation calculation module to achieve precise control of the sling tension.

[0033] Beneficial effects

[0034] Due to the above technical solution, compared with the prior art, the present invention has the following advantages and positive effects: By using the input variable to generate the universe scaling factor to dynamically adjust the range of the fuzzy universe in real time, when the deviation e is small, the universe is contracted, the width of the membership function graph is reduced, so that the number of control rules near the zero position of the error increases compared with before the variable universe, improving the utilization rate of the fuzzy rules. When the deviation e is large, the universe is expanded, the width of the membership function graph is increased, so that the fuzzy control rules are applicable to a larger gravity unloading range, thus enhancing the adaptive ability and real-time performance of the system. In addition, by introducing the desired force regulator, the gravity compensation value can be adjusted according to the actual working conditions, further improving the small-angle variable force control effect of the passive follow-up mechanism. The present invention can better meet the requirements of the space gravity unloading technology field, especially in the application scenario of active-passive hybrid following, showing excellent performance and stability. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] Figure 1 is the flowchart of the first embodiment of the present invention;

[0036] Figure 2 is the system structure diagram of the first and second embodiments of the present invention;

[0037] Figure 3 is the flow block diagram of the first and second embodiments of the present invention;

[0038] Figure 4 is the intermediate output variable ΔK in the flow block diagram of the first and second embodiments of the present invention P Membership function graph;

[0039] Figure 5 is the intermediate output variable ΔK in the first and second embodiments of the present invention P Schematic diagram of universe contraction. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0040] The following further elaborates the present invention in conjunction with specific embodiments. It should be understood that these embodiments are only used to illustrate the present invention and not to limit the scope of the present invention. In addition, it should be understood that after reading the content taught by the present invention, those skilled in the art can make various changes or modifications to the present invention, and these equivalent forms also fall within the scope defined by the appended claims of this application.

[0041] The first embodiment of the present invention relates to a control method for a suspended gravity compensation device, as Figure 1 shown, including the following steps:

[0042] Obtain the tension and inclination data of the suspension ropes at each moment;

[0043] Calculate the expected gravity compensation value based on the inclination angle of the suspension rope at the current moment, and calculate the deviation e between the suspension force and the expected gravity compensation value at the current moment;

[0044] Take the deviation e and the change rate ec of the deviation e with respect to time as the input variables of the fuzzy PID control algorithm, take the correction value of the PID control parameter as the intermediate output variable of the fuzzy PID control algorithm, and determine the basic universe of discourse of each of the input variables and the intermediate output variable;

[0045] According to the determined basic universe of discourse and the input variables, calculate the scaling factors of the universe of discourse of each input variable and the intermediate output variable at the corresponding moment, and dynamically adjust the universe of discourse range of the corresponding variable by using the calculated scaling factors of the universe of discourse;

[0046] Generate a PID control quantity based on the adjusted universe of discourse range.

[0047] The execution device of this embodiment is as Figure 2 shown, and this suspended gravity compensation system includes:

[0048] A force sensor for detecting the tension of the steel wire rope at each moment;

[0049] An angle sensor for detecting the inclination angle of the steel wire rope at each moment;

[0050] A controller for updating control parameters in real time according to the deployment state of the deployment mechanism, including a parameter setting module, a data acquisition module, a gravity compensation calculation module, and an expected force adjustment module, where:

[0051] The parameter setting module is used for communication configuration, setting of the gravity load F d , setting of the controller deviation threshold, and setting of relevant parameters for the ground microgravity simulation condition, such as deployment speed, suspension length, and load, etc.;

[0052] The data acquisition module is used for real-time acquisition of tension and inclination angle data;

[0053] The expected force adjustment module is used for calculating the expected gravity compensation value at the current moment according to the real-time acquired inclination angle data and the gravity load applied to the suspended gravity compensation;

[0054] The gravity compensation calculation module encapsulates the suspended gravity compensation control algorithm, and is used for calculating the control quantity according to the tension and inclination angle information of the steel wire rope and the expected gravity compensation value when the position servo mechanism carries the tension suspension mechanism and follows the spatial position of the deployment mechanism;

[0055] The motion control module is used for receiving the output control quantity of the compensation calculation module and driving the expansion and contraction of the electric cylinder to achieve precise control of the sling tension.

[0056] The flow framework of the suspension type gravity compensation control algorithm, that is, the variable universe fuzzy PID controller, is as follows Figure 3 shown, mainly including a fuzzy controller, a universe regulator, and a PID controller.

[0057] The input variables of the algorithm are set as: the vertical component F L of the wire rope tension F h and the difference from the expected gravity compensation value F d0 , denoted as the deviation e; the rate of change of the deviation e with respect to time, denoted as the rate of change ec.

[0058] In the fuzzy controller, a membership function composed of a trigonometric function and a Sigmoid function is used to fuzzify the input variables, and then defuzzification is performed through fuzzy inference and the centroid method to obtain the correction values ΔK P , K I , K D of the three control parameters P , ΔK I , ΔK D .

[0059] In the universe regulator, the range of the fuzzy universe is dynamically adjusted by multiplying the scaling factor by the boundary values of the universe, relatively increasing the number of control rules near the zero point of the error, and improving the utilization rate of the fuzzy rules.

[0060] In the PID controller, the output control quantity u(t), that is, the displacement adjustment amount Δz of the electric cylinder, is obtained through calculation.

[0061] The universe regulator embeds a calculation model of the scaling factor of the input variable universe and a calculation model of the scaling factor of the intermediate output variable universe. The input universe scaling factor model is of the proportional type, that is

[0062]

[0063] where E and EC are the boundary values of the basic universes of the input variables e and ec respectively; τ E , τ EC are the scaling factor coefficients; ε is a sufficiently small positive number approaching 0; the selection of the scaling factor coefficients follows the integral optimal criterion and is adjusted in real time according to the suspension force error and the error rate of change, and the specific form is taken as and

[0064] The output universe expansion factor model combines the monotonicity principle, that is, the proportional link ΔK P and the differential link ΔK DThe monotonicity should be consistent with the suspension force error, while the integral link ΔK I should have an opposite relationship with the suspension force error. At the same time, considering the respective tuning functions of the PID parameters, the following model is designed:

[0065]

[0066] where U KP 、U KI 、U KD are the boundary values of the basic domains of the output variables ΔK P 、ΔK I 、ΔK D respectively. Their lower limits are set to 0.003, 0.003, and 0.0001 to avoid position following failure; the parameters a i 、b i 、c i are tuned through experiments.

[0067] F d is the load gravity applied to the suspended gravity compensation device. The desired force regulator multiplies the inclination sensor signal θ by F d to obtain the dynamic desired force F d0 = F d (1 + sinθ), which further appropriately expands the fuzzy domain range of the variable domain fuzzy PID controller to improve the utilization rate of fuzzy rules and enhance the small-angle variable force control effect of the passive follow-up mechanism.

[0068] In addition, when the input variables e and ec both tend to decrease, the variable domain fuzzy PID controller will turn off the integral link to avoid integral saturation; otherwise, the integral link will be turned on.

[0069] The suspended gravity compensation control algorithm specifically includes the following steps:

[0070] Step 1: Determine the input variables e and ec of the control algorithm, as well as the intermediate output variables ΔK P 、ΔK I 、ΔK D and the output control quantity u(t); where the input variable e is the difference between the vertical component F L of the wire rope tension F h and the desired gravity compensation value F d0 , and the input variable ec represents the rate of change of the deviation e with respect to time; the intermediate output variables ΔK P 、ΔK I 、ΔK D are the PID control parameters K P 、K I 、K DThe correction value; the output control quantity u(t) is the displacement adjustment quantity Δz of the electric cylinder; further, the input variables e and ec are converted into the input variables ES and ECS after fuzzy processing. The basic domain of ES is [-10, 10], the basic domain of ECS is [-6, 6], and the basic domains of the intermediate output variables ΔK P 、ΔK I 、ΔK D are [-0.015, 0.015], [-0.00049, 0.00049], and [-0.014, 0.014] respectively;

[0071] Step 2, determine the domain scaling factors of the input variables and the domain scaling factor of the output variable embedded in the domain regulator. The domain scaling factors of the input variables e and ec are α E 、α EC , and their expressions are as follows:

[0072]

[0073] In the formula, E and EC are the boundary values of the basic domains of the input variables e and ec respectively; τ E 、τ EC are the scaling factor coefficients; ε is a sufficiently small positive number approaching 0; the selection of the scaling factor coefficients follows the integral optimal criterion and is adjusted in real time according to the error and the error change rate of the suspension force. The specific form is taken as and

[0074] The domain scaling factors of the intermediate output variables ΔK P 、ΔK I 、ΔK D are β KP 、β KI 、β KD , and their expressions are as follows:

[0075]

[0076] In the formula, U KP 、U KI 、U KD are the boundary values of the basic domains of the output variables ΔK P 、ΔK I 、ΔK D respectively. Their lower limits are set to 0.003, 0.003, and 0.0001 respectively to avoid position following failure; the parameters a i 、b i 、c i are determined by experiments.

[0077] Step 3: Determine that the quantization factors of the input variables \(e\) and \(ec\) are 0.3 and 0.5 respectively; the intermediate output variables \(\Delta K\) P \(\Delta K\) I \(\Delta K\) D have quantization factors of 200, 6250, and 214.286 respectively;

[0078] Step 4: Multiply the scaling factor by the boundary values of the basic universe of discourse, and use the quantization factor to generate the scaled fuzzy universe of discourse. Determine the quantization levels of the fuzzy universe of discourse. The quantization levels of the fuzzy universe of discourse are: negative large, negative medium, negative small, zero, positive small, positive medium, positive large, that is, the fuzzy sets are \(\{NB, NM, NS, ZE, PS, PM, PB\}\); among them, the membership function is composed of a triangular function and a sigmoid function.

[0079] Step 5: Design fuzzy rules in the form of "if-then" according to the quantization levels and membership functions determined in Step 4; since the monotonicity of the proportional link \(\Delta K\) P and the derivative link \(\Delta K\) D should be consistent with the suspension force error, while the monotonicity of the integral link \(\Delta K\) I should be opposite to the suspension force error, so two rule bases are established respectively; the fuzzy proportional and derivative control rule base is shown in Table 1, and the fuzzy integral control rule base is shown in Table 2.

[0080] Table 1 Fuzzy proportional and derivative control rule base

[0081]

[0082]

[0083] Table 2 Fuzzy integral control rule base

[0084]

[0085] Step 6: Through the defuzzification process of the fuzzy set using the centroid method, obtain a crisp value, which is used to correct the parameters of the PID controller, and according to the finally obtained corrected PID parameters, through calculate to obtain the output control amount \(u(t)\), that is, the displacement adjustment amount \(\Delta z\) of the electric cylinder.

[0086] Step 7: The motion control module receives the displacement adjustment amount \(\Delta z\) output by the PID controller, converts it into a control signal for the electric cylinder, and drives the electric cylinder to perform corresponding telescopic motion.

[0087] The specific process of determining the parameters of the output universe of discourse scaling factor model described in Step 2 above is as follows:

[0088] Step 1: Select a set of relevant parameters for the ground microgravity simulation conditions, including the deployment speed, suspension length, and load. Determine and set the intermediate output variable K according to experience. P , K I , K D Set the initial values of the three control parameters, start the gravity unloading test, and collect the signals of the force sensor and the inclination sensor in real time.

[0089] Step 2: Adjust the set value of the desired gravity compensation value F d0 . Optimize the intermediate control variables K P , K I , K D through the program so that the deviation e is controlled within 5% of the gravity compensation value F d0 .

[0090] Step 3: Change the relevant parameters of the ground microgravity simulation conditions, including the deployment speed, suspension length, and load. Repeat Steps 1 and 2 above.

[0091] Step 4: Use the linear autoregressive method to fit and obtain the coefficient matrix Y n

[0092]

[0093] The following further illustrates the action process of the universe regulator through specific embodiments.

[0094] Assume that ground microgravity simulation is carried out, and the working condition parameters are taken as a load of 30 kg, a deployment speed of 1 mm / s, and a suspension length of 1800 mm. Set the vertical component F L of the wire rope tension F h and the difference between the desired gravity compensation value F d0 is at most 1% of the load 30k, that is, 3 N.

[0095] Step 1: Determine that the basic universes of the input variables ES and ECS of the universe regulator are [-10, 10] and [-6, 6] respectively, and the basic universes of the intermediate output variables ΔK P , ΔK I , ΔK D are [-0.015, 0.015], [-0.00049, 0.00049], [-0.014, 0.014] respectively. Taking the intermediate output variable ΔK P as an example, the membership function graph under its initial universe is as shown in Figure 4 .

[0096] Step 2: In the universe regulator, determine the universe scaling factors of the embedded input variables and output variables, and set their initial values to 1.

[0097] Step 3, set the intermediate output variable K P , K I , K D Initial values of three control parameters, start the gravity unloading test, and collect the real-time signals F L (t i ) and θ(t i ). Assume that the input variables e and ec at time t i are 6 and 5.3 respectively.

[0098] Step 4, according to the input variables e and ec, the calculation formula of the domain scaling factor and the coefficient matrix Y n , obtain the scaling factors α i and α E of the input and output variables at time t EC , β KP , β KI , β KD are taken as (0.6, 0.53, 0.195104084, 9.66869024, 0.193811902) respectively. Accordingly, obtain the domain value of the intermediate output variable ΔK P as [-0.002927, 0.002927], and obtain the membership function graph of ΔK i at time t P as shown in Figure 5 .

[0099] Step 5, multiply the scaling factor by the boundary value of the fuzzy domain to determine the new fuzzy domain range of the intermediate output variable ΔK P . In this example, at time t i , the domain of the intermediate output variable K P is dynamically adjusted from [-0.015, 0.015] to [-0.0029, +0.0029], and the domain shrinks. As can be seen from Figure 5 , compared with before the variable domain, the number of rules in the same domain near the zero point increases, thereby improving the utilization rate of fuzzy rules. When the scaling factor is greater than 1, the domain will expand, the width of the membership function graph increases, and the fuzzy control rules are applicable to a larger gravity unloading range.

[0100] The second embodiment of the present invention relates to a control system of a suspended gravity compensation device, as shown in Figure 2 , including:

[0101] A force sensor for detecting the tension of the wire rope at each moment;

[0102] An angle sensor for detecting the inclination angle of the wire rope at each moment;

[0103] A controller, which is used to update control parameters in real time according to the deployment state of the deployment mechanism, including a parameter setting module, a data acquisition module, a gravity compensation calculation module, and an expected force adjustment module, where:

[0104] The parameter setting module is used to perform communication configuration, set the gravity load F d , set the controller deviation threshold, and set relevant parameters for the ground microgravity simulation working condition, such as deployment speed, suspension length, load, etc.;

[0105] The data acquisition module is used to collect tension and inclination data in real time;

[0106] The expected force adjustment module is used to calculate the expected gravity compensation value at the current moment according to the inclination data collected in real time and the gravity load applied to the suspended gravity compensation;

[0107] The gravity compensation calculation module encapsulates the suspended gravity compensation control algorithm. When the position servo mechanism carries the tension suspension mechanism and follows the spatial position of the deployment mechanism, it calculates the control quantity according to the wire rope tension and inclination information, as well as the expected gravity compensation value;

[0108] The motion control module is used to receive the output control quantity of the compensation calculation module and drive the telescoping of the electric cylinder to achieve precise control of the sling tension.

[0109] The process framework of the suspended gravity compensation control algorithm, that is, the variable universe fuzzy PID controller, is as Figure 3 shown, and mainly includes a fuzzy controller, a universe regulator, and a PID controller.

[0110] The input variables of the algorithm are set as: the vertical component F L of the wire rope tension F h and the difference between the expected gravity compensation value F d0 , denoted as the deviation e; the rate of change of the deviation e with respect to time, denoted as the rate of change ec.

[0111] In the fuzzy controller, a membership function composed of a trigonometric function and a Sigmoid function is used to perform fuzzy processing on the input variables. Subsequently, defuzzification is performed through fuzzy inference and the centroid method to obtain the correction values ΔK P 、K I 、K D of the three control parameters P 、ΔK I 、ΔK D .

[0112] In the universe regulator, by multiplying the scaling factor with the boundary values of the fuzzy universe, the range of the fuzzy universe is dynamically adjusted, relatively increasing the number of control rules near the zero-error position and improving the utilization rate of fuzzy rules.

[0113] In the PID controller, by calculating the output control quantity u(t), i.e., the displacement adjustment quantity Δz of the electric cylinder.

[0114] The universe regulator embeds a calculation model for the scaling factor of the input variable universe and a calculation model for the scaling factor of the intermediate output variable universe. The input universe scaling factor model is of the proportional type, i.e.,

[0115]

[0116] where E and EC are the boundary values of the basic universes of the input variables e and ec respectively; τ E and τ EC are the scaling factor coefficients; ε is a sufficiently small positive number approaching 0; the selection of the scaling factor coefficients follows the integral optimal criterion and is adjusted in real time according to the error and the rate of change of the error of the suspension force. The specific form is taken as and

[0117] The output universe expansion factor model combines the monotonicity principle, i.e., the monotonicity of the proportional link ΔK P and the differential link ΔK D should be consistent with the suspension force error, while the monotonicity of the integral link ΔK I should be in the opposite relationship with the suspension force error. At the same time, considering the respective tuning functions of the PID parameters, the following model is designed:

[0118]

[0119] where U KP 、U KI 、U KD are the boundary values of the basic universes of the output variables ΔK P 、ΔK I 、ΔK D respectively, and their lower limits are set to 0.003, 0.003, and 0.0001 respectively to avoid position following failure; the parameters a i 、b i 、c i are tuned through experiments.

[0120] F d is the load gravity applied to the suspended gravity compensation device. The desired force regulator multiplies the inclination sensor signal θ with F d to obtain the dynamic desired force F d0 = Fd (1 + sinθ) further expands the fuzzy domain range of the variable domain fuzzy PID controller appropriately to improve the utilization rate of fuzzy rules and enhance the small-angle variable force control effect of the passive servo mechanism.

[0121] In addition, when the input variables e and ec both tend to decrease, the variable domain fuzzy PID controller will turn off the integral link to avoid integral saturation; otherwise, it will turn on the integral link.

[0122] The suspended gravity compensation control algorithm specifically includes the following steps:

[0123] Step 1: Determine the input variables e and ec of the control algorithm, the intermediate output variables ΔK P , ΔK I , ΔK D and the output control quantity u(t); where the input variable e is the vertical component F L of the wire rope tension F h minus the expected gravity compensation value F d0 , and the input variable ec represents the rate of change of the deviation e with respect to time; the intermediate output variables ΔK P , ΔK I , ΔK D are the correction values of the PID control parameters K P , K I , K D ; the output control quantity u(t) is the displacement adjustment amount Δz of the electric cylinder; furthermore, after the input variables e and ec are fuzzified, they are converted into the input variables ES and ECS. The basic domain of ES is [-10, 10], the basic domain of ECS is [-6, 6], and the basic domains of the intermediate output variables ΔK P , ΔK I , ΔK D are [-0.015, 0.015], [-0.00049, 0.00049], [-0.014, 0.014] respectively;

[0124] Step 2: Determine the input variable domain scaling factors and the output variable domain scaling factors embedded in the domain regulator, where the domain scaling factors of the input variables e and ec are α E , α EC , and their expressions are as follows:

[0125]

[0126] In the formula, E and EC are the boundary values of the basic domains of the input variables e and ec respectively; τ E , τ ECis the scaling factor coefficient; ε is a sufficiently small positive number approaching 0; the selection of the scaling factor coefficient follows the integral optimal criterion and is adjusted in real time according to the error and error change rate of the suspension force, and its specific form is taken as and

[0127] The scaling factors of the intermediate output variables ΔK P 、ΔK I 、ΔK D are β KP 、β KI 、β KD , and their expressions are as follows:

[0128]

[0129] In the formula, U KP 、U KI 、U KD are the boundary values of the basic universes of the output variables ΔK P 、ΔK I 、ΔK D respectively, and their lower limits are set to 0.003, 0.003, 0.0001 respectively to avoid position following failure; the parameters a i 、b i 、c i are determined by experiments.

[0130] Step 3, determine that the quantization factors of the input variables e and ec are 0.3 and 0.5 respectively; the quantization factors of the intermediate output variables ΔK P 、ΔK I 、ΔK D are 200, 6250, 214.286 respectively;

[0131] Step 4, multiply the scaling factor by the boundary value of the basic universe and use the quantization factor to generate the scaled fuzzy universe. Determine the quantization levels of the fuzzy universe, and the quantization levels of the fuzzy universe are: negative large, negative medium, negative small, zero, positive small, positive medium, positive large, that is, the fuzzy sets are {NB, NM, NS, ZE, PS, PM, PB}; among them, determine that the membership function is a combination of a triangular function and a Sigmoid function.

[0132] Step 5, design fuzzy rules in the form of "if-then" according to the quantization levels and membership functions determined in Step 4; since the monotonicity of the proportional link ΔK P and the differential link ΔK D should be consistent with the suspension force error, while the integral link ΔK IThe monotonicity should have an opposite relationship with the suspension force error, so two rule bases are established respectively; the fuzzy proportional and differential control rule base is shown in Table 1, and the fuzzy integral control rule base is shown in Table 2.

[0133] Table 1 Fuzzy proportional and differential control rule base

[0134]

[0135]

[0136] Table 2 Fuzzy integral control rule base

[0137]

[0138] Step 6, by using the centroid method to defuzzify the fuzzy set, a crisp value is obtained, which is used to correct the parameters of the PID controller. According to the finally obtained corrected PID parameters, through calculation, the output control quantity u(t) is obtained, that is, the displacement adjustment amount Δz of the electric cylinder.

[0139] Step 7, the motion control module receives the displacement adjustment amount Δz output by the PID controller, converts it into the control signal of the electric cylinder, and drives the electric cylinder to perform corresponding telescopic motion.

[0140] The specific process for determining the parameters of the output domain scaling factor model described in Step 2 above is as follows:

[0141] Step 1, select a set of relevant parameters of the ground microgravity simulation working conditions, including the deployment speed, suspension length and load. Determine and set the initial values of the three control parameters K P 、K I 、K D according to experience, start the gravity unloading test, and collect the signals of the force sensor and the tilt sensor in real time;

[0142] Step 2, adjust the set value of the desired gravity compensation value F d0 , and optimize the intermediate control variables K P 、K I 、K D through the program, so that the deviation e is controlled within 5% of the gravity compensation value F d0 .

[0143] Step 3, change the relevant parameters of the ground microgravity simulation working conditions, including the deployment speed, suspension length and load. Repeat the above Steps 1 and 2.

[0144] Step 4, use the linear autoregressive method to fit and obtain the coefficient matrix Y n

[0145]

[0146] The working process of the domain regulator is further explained below through a specific embodiment.

[0147] Assuming that the ground microgravity simulation is carried out, the working parameters are set as load 30kg, deployment speed 1mm / s, and suspension length 1800mm. Set the wire rope tension F L The vertical component F h and the desired gravity compensation value F d0 The maximum difference is no more than 1% of the load 30k, i.e. 3N.

[0148] Step 1: Determine the basic domains of the input variables ES and ECS of the domain regulator to be [-10, 10] and [-6, 6] respectively, and the intermediate output variable ΔK P , ΔK I , ΔK D The basic domains are [-0.015, 0.015], [-0.00049, 0.00049], [-0.014, 0.014]. P Taking the initial domain as an example, the membership function diagram is as follows: Figure 4 shown.

[0149] Step 2: In the domain regulator, determine the domain scaling factors of the embedded input variables and output variables, and set their initial values ​​to 1.

[0150] Step 3: Set the intermediate output variable K P , K I , K D Initial values ​​of the three control parameters, start the gravity unloading test, and collect the real-time signals F of the force sensor and the inclination sensor L (t i ) and θ(t i ). Assume that we calculate t i The input variables e and ec at time t are 6 and 5.3 respectively.

[0151] Step 4: Calculate the formula and coefficient matrix Y based on the input variables e and ec domain scaling factor n , we get t i Input and output variable scaling factor α at the moment E , α EC , β KP , β KI , β KD They are respectively taken as (0.6, 0.53, 0.195104084, 9.66869024, 0.193811902). Based on this, the intermediate output variable ΔK is obtainedP The universe of discourse takes values in [-0.002927, 0.002927], and t is obtained i At time t, ΔK P The membership function graph of Figure 5 is shown as follows.

[0152] Step 5: Multiply the scaling factor by the boundary values of the fuzzy universe of discourse to determine the intermediate output variable ΔK P The new fuzzy universe of discourse range. In this example, at time t i , the universe of discourse of the intermediate output variable K P dynamically adjusts from [-0.015, 0.015] to [-0.0029, +0.0029], and the universe of discourse shrinks. As Figure 5 can be seen, compared with before the variable universe of discourse, the number of rules under the same universe of discourse near zero increases, thus improving the utilization rate of fuzzy rules. When the scaling factor is greater than 1, the universe of discourse will expand, the width of the membership function graph increases, and the fuzzy control rules are applicable to a larger gravity unloading range.

Claims

1. A control method for a suspended gravity compensation device, characterized in that: The following steps are involved: Obtain the tension and inclination data of the suspension rope at each moment; Calculate the expected gravity compensation value based on the inclination angle of the suspension rope at the current moment, and calculate the deviation e between the suspension force and the expected gravity compensation value at the current moment; Taking the deviation e and the rate of change ec of the deviation e with respect to time as input variables of the fuzzy PID control algorithm, taking the correction value of the PID control parameter as the intermediate output variable of the fuzzy PID control algorithm, and determining the basic domain of each of the input variables and the intermediate output variable; According to the determined basic domain and the input variable, the domain scaling factor of each input variable and the intermediate output variable at the corresponding time is calculated respectively, and the domain range of the corresponding variable is dynamically adjusted by using the calculated domain scaling factor; Generate PID control variables based on the adjusted domain range.

2. The method according to claim 1, characterized in that The domain scaling factor of the input variable is calculated by the following formula: Among them, α E and α EC are the domain expansion factors of input variables e and ec, respectively. E and EC are the boundary values ​​of the basic domain of input variables e and ec, respectively. E and τ EC are the expansion factor coefficients of input variables e and ec respectively, and ε is a positive number tending to 0.

3. The method according to claim 2, characterized in that The scaling factor coefficients are selected based on an integral optimality criterion.

4. The method according to claim 1, characterized in that: The domain scaling factor of the intermediate output variable is calculated by the following formula: Among them, β KP , β KI , β KD They are the output variables ΔK P , output variable ΔK I , output variable ΔK D The universe expansion factor, U KP , U KI , U KD They are the output variables ΔK P , output variable ΔK I , output variable ΔK D The boundary value of the basic domain of i , b i and c i They are all setting parameters, and i=1, 2, 3.

5. The method according to claim 4, characterized in that The setting parameters are obtained by fitting multiple groups of known operating parameters of the suspended gravity compensation device through a linear autoregressive method.

6. The method according to claim 1, characterized in that The desired gravity compensation value is calculated by the following formula: F d0 =F d (1+sinθ) Among them, F d0 is the expected gravity compensation value, F d is the load gravity applied to the suspended gravity compensation device, and θ is the inclination angle of the suspension rope.

7. The method according to claim 1, characterized in that The suspension force is the vertical component of the rope tension.

8. The method according to claim 1, characterized in that: The generating of the PID control quantity based on the adjusted domain range includes: Based on the adjusted domain range, the first rule base for the proportional link and the differential link, and the second rule base for the integral link are respectively constructed according to the set quantization level and membership function, and fuzzy reasoning is performed; The fuzzy set is defuzzified using the centroid method to obtain the clarified intermediate output variable, and then the PID control variable is generated using the modified PID control parameter.

9. A control system for a suspended gravity compensation device, characterized in that: include: Force sensor, used to detect the tension of the suspension rope; Angle sensor, used to detect the inclination of the suspension rope; The controller includes a parameter setting module, a data acquisition module, a gravity compensation calculation module and a desired force adjustment module, wherein: Data acquisition module, used to collect tension and inclination data; The expected force adjustment module is used to calculate the expected gravity compensation value at the current moment according to the inclination data collected in real time and the gravity load applied to the suspended gravity compensation; A gravity compensation calculation module, used to execute any method as claimed in claims 1-8; The motion control module is used to drive the extension and retraction of the electric cylinder according to the output control quantity of the compensation calculation module, so as to realize the precise control of the tension of the sling.

Images