A multi-point synchronous loading device and control method for solar wing ground assembly

Abstract

The application aims to provide a multi-point synchronous loading device and control method for solar wing ground assembly, solve the tension and compression assembly load required by multiple assembly points of circular solar wing ground assembly, and is composed of a supporting frame and eight loading modules; the supporting frame comprises a horizontal frame and an inverted frame, lifting feet are respectively installed at four corners of the horizontal frame, the height of the lifting feet is adjustable, and the horizontal frame is supported by the lifting feet; the side vertical frame is perpendicular to the horizontal frame and is fixedly installed on a side cross bar of the horizontal frame, multiple quick mounting plates are installed at different positions on the side vertical frame of the supporting frame, and the loading modules are mounted on the quick mounting plates.

Description

Technical Field

[0001] This invention belongs to the technical fields of aerospace device assembly, mechanical engineering and control engineering, and specifically relates to a multi-point synchronous loading device and control method for ground assembly of solar panels. Background Technology

[0002] my country's aerospace engineering is developing rapidly. Circular solar arrays, as energy sources for spacecraft, possess advantages such as light weight and high packing ratio, meeting the ever-increasing energy supply requirements and the need for miniaturization and lightweight design of spacecraft. The clamping assembly of circular solar arrays places high demands on their own assembly stress; improper stress control can cause the solar array to fail to deploy or even be damaged. The clamping assembly process requires applying tensile and compressive loads to multiple assembly points simultaneously. However, due to the complex mechanical properties of solar arrays and the need for precise control of the loading force at the assembly points, the control method must meet high precision, high disturbance rejection, and high synchronization requirements. There is an urgent need for equipment and control methods for ground-based clamping assembly of circular solar arrays. Currently, there is no corresponding equipment or technology to solve these assembly problems. Summary of the Invention

[0003] Based on the above-mentioned technological status, the purpose of this invention is to provide a multi-point synchronous loading device and control method for ground assembly of solar panels, so as to solve the tensile and compressive assembly loads required for multiple assembly points of circular solar panels.

[0004] The technical solution adopted in this invention is as follows: a multi-point synchronous loading device and control method for ground assembly of solar panels includes a mechanical system and a control method.

[0005] The mechanical system includes a loading module and a support frame.

[0006] The loading module includes a loading device and a flexible six-degree-of-freedom adjustment device. The loading device is driven by a motor, reducer, and lead screw system. A mechanical interface is installed at the output end of the lead screw system to apply tensile and compressive loads to the solar array assembly points. A force sensor is installed inside the mechanical interface to measure the load value in real time. The flexible six-degree-of-freedom adjustment device is installed below the loading device and includes a spring bracket, a three-degree-of-freedom slide, and a quick-mount plate. The spring bracket is supported by four spring columns. The length of the elastic columns can be adjusted by turning the nuts on the columns, thereby adjusting the three-degree-of-freedom installation space at the top of the spring bracket. The three-degree-of-freedom slide is mounted on the lower end of the spring bracket, and together with the spring bracket, it realizes the six-degree-of-freedom attitude adjustment of the loading device; the quick-installation plate includes a plate body and a quick-clamping bolt; the plate body is designed with an I-shaped groove that mates with the bottom of the three-degree-of-freedom slide for positioning; the quick-clamping bolt includes a hand-tightening nut, a strip pressure plate, a flexible washer, a bolt, and a cam handle. After the bottom of the three-degree-of-freedom slide body is positioned in conjunction with the I-shaped groove in the plate body, the strip pressure plate is rotated 90 degrees, and the hand-tightening nut is driven by turning the cam handle to clamp the strip pressure plate, thus achieving quick and fixed assembly of the two.

[0007] The support frame includes a main frame, lifting feet, and sliding feet. The main frame is an L-shaped symmetrical frame structure. The lifting feet are installed at the bottom of the main frame and the posture of the main frame can be adjusted by adjusting the support height through a screw. The sliding feet are equipped with universal pulleys to facilitate the movement of the support frame.

[0008] The solar array has two mirror symmetrical assembly modes, ±Y, and the assembly space is small. The support frame adopts a symmetrical frame structure and has corresponding quick-mount plates pre-installed in a mirror symmetrical manner. When switching between the two mirror symmetrical assembly modes of the solar array, it is only necessary to assemble the loading device and the flexible six-degree-of-freedom adjustment device in the loading module as a whole unit on the quick-mount plate in the mirror position, which greatly saves the device adjustment time and adjustment space.

[0009] The control method includes control system hardware, control method for a single loading device, control method for multiple loading devices, and a tiered loading strategy.

[0010] The control system hardware consists of a host computer, a motion controller, an analog signal acquisition module, a servo motor driver, a servo motor, and an S-shaped force sensor. For each loading device in this invention, the S-shaped force sensor is responsible for inputting the current loading force as a voltage value into the analog signal acquisition module. The analog signal acquisition module converts this voltage signal into a digital signal and feeds it back to the host computer via Ethernet through the EtherCAT bus between itself and the motion controller. Subsequently, the host computer converts the current digital signal into a force signal, compares this force signal with the set target loading force, and calculates the motor speed signal through the control methods for a single loading device and multiple devices. This motor speed signal is transmitted to the motion controller via Ethernet, and then the motion controller transmits it to the corresponding motor's servo driver via EtherCAT. The servo driver executes the signal to drive the motor.

[0011] The control method for the single loading device includes an adaptive PID control algorithm based on recursive least squares with a forgetting factor. Because the solar array has a flexible internal structure, the deformation of this structure during the application of force causes a sudden change in the internal stiffness of the solar array, resulting in the overall stiffness of the solar array becoming a time-varying variable.

[0012] K→K(k) (Discrete form)

[0013] Linear time-invariant control systems struggle to adapt to such abrupt changes in stiffness, leading to performance variations and, in severe cases, oscillations and convergence failure. The mathematical formula for the adaptive PID control algorithm based on recursive least squares with a forgetting factor is described below:

[0014]

[0015] In the formula, F i0 F represents the desired current applied force. i (k) represents the magnitude of the actual applied force fed back by the tension sensor, e i (k) represents the error term of the feedback control system. K P T is the proportional gain of the PID controller. I Let T be the integration time constant. D Let T be the differential time constant, T be the system sampling period, and k be the number of samples. M is a constant for the identified system stiffness.

[0016] The control system identifies the system stiffness parameters online using a recursive least squares method with a forgetting factor, and then corrects the PID parameters of each loading stage in real time based on the system stiffness. This allows the control system to adapt to the sudden stiffness changes of the solar array caused by the stress on the flexible structure, thereby ensuring the loading accuracy and synchronization of the system. The identification process of the recursive least squares method with a forgetting factor is as follows: (1) Real-time acquisition of encoder feedback information from the motor during loading to calculate the current rotation angle of the motor; (2) Real-time acquisition of the loading force during loading; (3) Identification of the overall stiffness of the system based on the acquired encoder feedback information and the loading force. The mathematical formula of this method is described as follows:

[0017] For this loading system, assuming that at the k-th output signal sampling, the measured motor rotation angle is θ(k), the measured loading force is F(k), and the parameter to be identified is the system stiffness. If each sampling moment during the loading process is considered quasi-static, then after the kth sampling, the motor rotation angle θ(k), the loading force F(k), and the identified system stiffness are... The following relationship exists between them:

[0018]

[0019]

[0020] In the formula, e(k) represents the actual sampled loading force F(k) and the identified system stiffness at the k-th output signal sampling time. The predicted loading force value calculated from the motor rotation angle θ(k) The error between them. Assuming that the output variable F(k) and the observable variable θ(k) are observed k times at times i = 1, 2, ..., k, and the error of each observation is e(1), e(2), ..., e(k), then k linear equations can be established, which can be expressed in matrix form as follows:

[0021]

[0022]

[0023] The square of the sum of all identification errors in k iterations is defined as the evaluation function of the recursive least squares algorithm with a forgetting factor, which can be expressed as follows:

[0024]

[0025] The least squares estimate of the parameters corresponds to the parameters when the objective function J(k) reaches its minimum value after the k-th sampling. Let J be Find the first derivative and set it to 0, then

[0026]

[0027] Based on the above formula, the parameters after the kth sampling can be obtained. Least squares estimator:

[0028]

[0029] For online real-time parameter identification, it is necessary to... The estimation formula is transformed into a recursive algorithm, thus obtaining the recursive least squares algorithm. However, what we need to identify is the system stiffness at a certain time k. We only need the sampled values ​​near time k, instead of all historical sampled values. Therefore, we can use least squares with a forgetting factor to reduce the influence of historical data on the current time. Introducing the vectors P(k) and γ(k), we obtain the formula for the recursive least squares method with a forgetting factor:

[0030]

[0031] In the formula, λ is the forgetting factor, representing the weight of the data. By appropriately selecting the value of λ, the current time can be obtained. The optimal parameter identification value, and then based on To correct the K of the PID controller in real time p This achieves an adaptive effect.

[0032] The control method for the multiple loading devices is a deviation coupling control algorithm based on average values. Its implementation principle is to sum and average the deviations between the target force value and the current force value at each clamping point.

[0033]

[0034] Subsequently, the difference between the error at the current point and this average value is corrected by a PID controller and then passed to the deviation calculation of each loading point. The new error for each loading point is expressed as:

[0035]

[0036] In the formula, K′ P To correct the proportional gain of the PID controller for error, T′ I T′ is the integral time constant of the error-corrected PID controller. DThis is the derivative time constant of the PID controller for error correction. In other words, this error is fed back to the single-point adaptive PID controller based on recursive least squares, thereby achieving "information sharing" among all loading points. Each point corrects itself using the average error of all loading points as the target value, thus achieving high-precision synchronous control. The final control algorithm formula, combining the control methods for a single loading device and those for multiple loading devices, is as follows:

[0037]

[0038] Because the solar array experiences sudden stiffness changes due to flexible deformation during loading, and these changes have varying impacts on each loading point—meaning the stiffness change, duration of impact, and timing differ—it's difficult to guarantee synchronization if the control method for each loading point operates independently. Therefore, a synchronization control method is needed. This control algorithm incorporates current information from other points into the controller at each point, allowing each point to adjust its control input based on the average value of all points. This minimizes synchronization errors across multiple points and improves synchronization.

[0039] The stepped loading strategy involves dividing the target loading force into segments, with the loading force increasing in stages until the final target force value is reached. This stepped loading strategy is adopted because direct loading has two drawbacks: First, for feedback control systems, the convergence curve before asymptotic stability approximates an exponential curve, meaning that the convergence speed is fast when the error is large and slow when the error is small. When the target force is significantly larger than the current force, a large force will be applied to the solar array within a short timeframe during loading, which can impact the internal structure of the solar array and potentially damage it. Therefore, in the loading strategy, when the difference between the target force and the current force (i.e., the force to be loaded) exceeds a certain threshold, the loading process is segmented, with each segment loading only a portion of the force instead of the entire force. This reduces the impact of the impulse caused by the loading force on the solar array. Second, due to frictional resistance and the complex mechanical properties of the solar array, when the target force is significantly larger than the current force, it is difficult to ensure the synchronization of the loading forces when multiple motors are simultaneously loading. Segmented loading breaks down large loads into smaller ones, helping to ensure the synchronization of loading across multiple motors and reducing the impact of uneven loading on the solar panels. Furthermore, at startup, we use 60% of the target load for each segment. This is done to allow the motors to overcome certain frictional resistance during startup, ensuring the synchronization of load forces throughout the loading process. The specific execution process is as follows:

[0040] (1) First, set the final target force and the maximum loading force for each loading step, and check if the parameters are set correctly. If they are set correctly, proceed to the next step. (2) During the initial loading, convert the analog signal fed back by the S-type tension sensor on each loading device at the current moment into a force value and take the average value. Based on the difference between the final target force and the average current force value, compare this difference with 60% of the maximum loading force in a single step. If this difference is greater than 60% of the maximum loading force in a single step, set the current target force to the current force value + 60% of the maximum loading force in a single step; if this difference is less than 60% of the maximum loading force in a single step, set the current target force as the final target force. (3) After determining the current target force for the next loading, the host computer software inputs the current loading force in the form of voltage to the analog signal acquisition module through the S-type tension sensor of each loading device. The analog signal acquisition module converts the voltage signal into a digital signal and feeds the digital signal back to the host computer via Ethernet through the EtherCAT bus between the module and the motion controller. Subsequently, the host computer converts the current digital quantity into a force signal, compares the magnitude of this force signal with the preset target loading force, and calculates the motor speed signal through the control methods of a single loading device and multiple devices. This motor speed signal is transmitted to the motion controller via Ethernet, and then the motion controller transmits it to the corresponding motor servo driver via EtherCAT. The servo driver executes the signal to drive the motor to move. The motor movement is converted into the displacement of the steel belt on the loading device, thereby loading the force on the clamping and releasing device on the solar panel. (4) After each loading step is completed, the system determines whether the current target force is equal to the final target force. If the current target force is not equal to the final target force, (2) to (4) are repeated. In (2), "When loading for the first time, convert the analog quantity fed back by the S-type tension sensor on each loading device at the current moment into a force value and take the average value, and according to the difference between the final target force and the average value of the current force" is replaced with "according to the difference between the final target force and the current target force". In (2), all "60% of the maximum loading force in a single step" is replaced with "the maximum loading force in a single step". If the current target force is equal to the final target force, the current loading is ended.

[0041] The advantages of the technical solution of this invention are:

[0042] 1. The support frame of this invention adopts a symmetrical frame structure, which saves the adjustment time and space required for the device to adapt to mirror assembly conditions;

[0043] 2. In this invention, a quick-installation plate is designed between the loading module and the support frame, which greatly saves the preparation time before the equipment is put into operation;

[0044] 3. The bottom of the support frame of the device is designed with lifting feet to adjust the overall posture of the equipment and reduce assembly errors;

[0045] 4. This invention combines an adaptive PID algorithm based on recursive least squares, a deviation coupling control algorithm based on average value, and a stepped loading strategy. This avoids oscillations and divergences caused by changes in the control system structure due to sudden changes in the internal stiffness of the solar array, ensures the consistency of loading force among multiple points during the loading process, and reduces the impact of the loading force on the solar array. Attached Figure Description

[0046] Figure 1 This is a schematic diagram of the overall structure of the multi-point synchronous loading device of the present invention;

[0047] Figure 2 This is a schematic diagram of the +Y working condition configuration of the present invention;

[0048] Figure 3 This is a schematic diagram of the Y-condition configuration of the present invention;

[0049] Figure 4 This is a schematic diagram of the loading module structure of the present invention;

[0050] Figure 5 This is an exploded view of the loading module of the present invention;

[0051] Figure 6 This is a schematic diagram of the quick-installation plate structure of the present invention;

[0052] Figure 7 This is a schematic diagram of the control method for a single loading module of the present invention;

[0053] Figure 8 This is a schematic diagram of the control method for multiple loading modules of the present invention;

[0054] Figure 9 This is a schematic diagram of the stepped loading strategy of the present invention;

[0055] Figure 10 This is the online identification result of the total system stiffness of the single loading module control method of the present invention when the recursive least squares method is applied to the actual solar array from a load of 500N to 4000N.

[0056] Figure 11 This is a graph showing the actual loading curve on the solar panel when the maximum single-step loading force is 200N, from 300N to 1000N, and the maximum deviation curve between each point.

[0057] Figure 12 This is a graph showing the actual loading curve on the solar panel when the maximum single-step loading force is 200N, from 1000N to 2000N, and the maximum deviation curve between each point.

[0058] Figure 13 This is a graph showing the actual loading curve of the solar array from 2000N to 3000N when the maximum single-step loading force is 200N, and the maximum deviation curve between each point.

[0059] Figure 14 This is a graph showing the actual loading curve of the solar array when the maximum single-step loading force is 200N, from 3200N to 4000N, and the maximum deviation curve between each point.

[0060] In the diagram: 1. Loading module, 2. Support frame, 3. Quick-install plate, 4. Lifting foot, 5. Sliding foot, 6. Loading device, 7. Spring bracket, 8. Three-degree-of-freedom slide, 9. Mounting bolt, 10. Hand-tightening nut, 11. Strip pressure plate, 12. Plate body, 13. Flexible gasket, 14. Stud, 15. Cam handle. Detailed Implementation

[0061] Figure 1 This is a schematic diagram of the overall structure of the multi-point synchronous loading device of the present invention. The multi-point synchronous loading device of the present invention for ground assembly of solar array consists of a support frame 2 and eight loading modules 1.

[0062] The support frame 2 includes a horizontal frame and an inverted frame. The horizontal frame has adjustable feet 4 at its four corners to support it. Multiple sliding feet 5 with rollers are also installed on the bottom surface of the horizontal frame to move the entire support frame 2. The side frame is perpendicular to the horizontal frame and fixedly installed on a horizontal bar on one side of the horizontal frame. Eleven quick-installation plates 3 are installed at different positions on the side frame of the support frame 2 for mounting... Figure 4 As shown in the loading module 1, specifically, the quick-installation plate 3 is fixedly mounted on the support frame 2 using mounting bolts 9. On the support frame 2, the multiple quick-installation plates 3 are distributed in two forms: +Y assembly configuration and -Y assembly configuration, as shown below. Figure 2 The diagram shows the +Y configuration of the present invention. Two reinforcing support rods are fixedly installed inside the side frame. The two reinforcing support rods are symmetrically distributed and are inclinedly connected between the upper and lower crossbars of the side frame. In the +Y assembly condition, four loading modules 1 are installed on the upper crossbar of the side frame, and three loading modules are installed on the lower crossbar of the side frame. The installation positions of each loading module 1 correspond to the positions of the clamping and releasing devices on the solar panel in the +Y assembly condition.

[0063] Figure 3This is a schematic diagram of the -Y working condition configuration of the present invention. In this working condition, the installation positions of the eight loading modules on the side frame are mirror images of the installation positions in the +Y assembly condition.

[0064] See Figure 4 This is a schematic diagram of the loading module structure of the present invention, and its exploded view is shown below. Figure 5 As shown. The loading module 1 includes a loading device 6, a spring bracket 7, and a three-degree-of-freedom slide table 8. The loading device 6 consists of a servo motor, a harmonic reducer, a lead screw drive module, an output shaft, an anti-detachment clamp, a support frame, a locking nut, and an adaptive hook. The loading device 1 is an overall shaft structure, installed above the spring bracket 7, and its connection to the pressure release device on the solar panel is fixed through an interface connection module to realize the loading action. The servo motor drives the lead screw drive module to rotate through the harmonic reducer. The lead screw drive module drives the lead screw nut to perform linear feed motion through the rotation of the lead screw, thereby driving the output shaft fixedly connected to the lead screw nut to perform linear feed motion. The output shaft of the lead screw drive module drives the adaptive hook to move linearly back and forth. The adaptive hook includes a connecting contact angle, a connecting rod one, an intermediate connector, an end connector, and a connecting rod two. The front end of the connecting contact angle has a hook-shaped structure, and the rear end of the connecting contact angle has a hinged structure. Both connecting rod one and connecting rod two have three through holes in the middle. The distance between the two through holes on both sides and the central through hole is the same. The connecting contact angles are arranged in pairs. The hinged structure at the rear end of the two connecting contact angles is hinged and installed at both ends of connecting rod one using screws. The middle hole of connecting rod one is hinged and installed at the front hinge of the intermediate connector using screws. The intermediate connector is a hinge structure composed of mutually perpendicular front hinges and rear hinges. The structure composed of the connecting contact angle, connecting rod one, and intermediate connector is arranged symmetrically in two sets. The rear hinges of the two intermediate connectors are hinged and installed at the two through holes on the side of connecting rod two using screws. The middle through hole of connecting rod two is hinged and installed at the front end of the end connector using screws. The end connector shown includes a connector interface, a double-threaded sleeve, a connector bolt, and a threaded sleeve. One end face of the connector interface is provided with a hinge that is hinged to the intermediate hole of the connecting rod, and the other end face forms a threaded blind hole. The double-threaded sleeve is installed in the threaded blind hole of the connector interface through a threaded pair. The connector bolt is installed inside the double-threaded sleeve. The strength of the connector bolt is greater than the strength of the inner sleeve of the double-threaded sleeve. The threaded sleeve is installed at the end of the connector bolt. The threaded sleeve is connected to the tension sensor through a threaded pair.

[0065] The spring bracket 7 includes a stud, a V-shaped plate, a second locking nut, an adjusting nut, a spring, a spring sleeve, a base plate, a bottom bracket, and an anti-loosening nut. The stud is installed above the elongated hole of the bottom bracket. The bottom bracket includes two side-by-side components, each with an elongated hole on each side. A set of gravity compensation components is installed in each elongated hole. Each set of gravity compensation components includes a stud, a V-shaped plate, a second locking nut, an adjusting nut, a spring, and a spring sleeve. The V-shaped plate, the second locking nut, the adjusting nut, the spring, and the spring sleeve pass through the stud from top to bottom and are installed above the elongated hole of the bottom bracket. There are two V-shaped plates, which are fixedly installed on both sides above the base plate. A stud passes through each end of the V-shaped plate. The second locking nut is used to lock the V-shaped plate and prevent it from sliding up and down relative to the stud. The compression of the spring can be adjusted by adjusting the adjusting nut, thereby achieving three-degree-of-freedom rotation adjustment at the top. At the same time, because the spring has the ability to deform, the loading device 6 has a certain self-adaptive ability to installation errors, reducing contact stiffness and installation stress. The three-degree-of-freedom slide table 8 is mounted on the quick-installation plate 3. The three-degree-of-freedom slide table 8 includes a scissor lift, slide table one, and slide table two. The scissor lift is mounted on top of slide table one, and slide table one is mounted on top of slide table two. Slide table one is horizontally and vertically mounted on slide table two. The two slide tables together can achieve ±50mm displacement adjustment along the horizontal XY axis, and the scissor lift can achieve ±10mm micro-adjustment in the vertical direction, providing three-degree-of-freedom translational micro-adjustment for the loading device 6. Through series mounting with the spring bracket, the fine-tuning mechanism can achieve six-degree-of-freedom error compensation for the assembly position of the loading device 6. The spring bracket 7 and the three-degree-of-freedom slide table 8 together provide six degrees of freedom for the loading device, allowing the loading device 6 to adaptively adjust during loading to ensure its central axis is coaxial with the pressure release device on the solar panel.

[0066] Figure 6 This is a schematic diagram of the quick-installation plate structure of the present invention. The quick-installation plate 3 includes a hand-tightening nut 10, a strip pressure plate 11, a plate body 12, a flexible gasket 13, a bolt 14, and a cam handle 15. The surface of the plate body 12 is provided with an I-shaped groove. During installation, the three-degree-of-freedom slide 8 is positioned in the I-shaped groove of the plate body 12. Then, the strip pressure plate 11 is rotated 90 degrees so that it engages with the positioning groove on the bottom periphery of the three-degree-of-freedom slide 8. By turning the cam handle 15, the bolt 14 is pulled axially and the hand-tightening nut 10 is driven to press the strip pressure plate 11 tightly, thereby realizing the quick fixed assembly of the three-degree-of-freedom slide and the installation plate.

[0067] In use, the support frame needs to be moved near the solar panel via the sliding foot 5, and then the lifting foot 4 is lowered to fix the support frame 1 and adjust it to be level. Subsequently, depending on whether the solar panel is in a +Y or -Y assembly configuration, the quick-mount plate 3 is installed on the corresponding position of the support frame 1, and then the loading device and quick-mount plate 1 are assembled as follows. Figure 4 As shown. At this point, the invention is fully assembled. By adjusting the three-degree-of-freedom slide 8 and the spring bracket 7, the central axis of the loading device 6 is made coaxial with the clamping and releasing device on the solar panel. Then, by adjusting the three-degree-of-freedom slide 8, the loading device 6 is connected to and fixed with the clamping and releasing device on the solar panel. At this point, the loading system is fully assembled and can be loaded.

[0068] For the purposes of this invention, the loading step refers to the entire process of decomposing the final target force into multiple continuous, monotonically increasing or monotonically decreasing current target forces according to a step-by-step loading strategy, and then loading from the current force to the current target force each time.

[0069] The eight loading devices 6 of this invention, together with the solar array, form a complete loading system. The loading process after adopting a stepped loading strategy is as follows: Figure 9As shown, the specific description is as follows: (1) First, set the magnitude of the final target force and the maximum loading force of each loading step, and check whether the parameters are set correctly. If the parameters are set correctly, proceed to the next step. (2) During the initial loading, convert the analog quantity fed back by the S-type tension sensor of each loading device 6 at the current moment into a force value and take the average value. Based on the difference between the final target force and the average value of the current force, compare this difference with 60% of the maximum loading force of a single step. If this difference is greater than 60% of the maximum loading force of a single step, then set the current target force to the current force value + 60% of the maximum loading force of a single step; if this difference is less than 60% of the maximum loading force of a single step, then set the current target force as the final target force. (3) After determining the magnitude of the current target force for the next loading, the host computer software inputs the current loading force magnitude in the form of voltage into the analog quantity acquisition module through the S-type tension sensor of each loading device 6. The analog quantity acquisition module converts the voltage signal into a digital signal and feeds the digital signal back to the host computer through the EtherCAT bus between it and the motion controller via Ethernet. Subsequently, the host computer converts the current digital quantity into a force signal, compares the magnitude of this force signal with the preset target loading force, and generates a motor speed signal through calculations using the control methods of a single loading device and multiple devices. This motor speed signal is transmitted to the motion controller via Ethernet, and then transmitted by the motion controller to the corresponding motor servo driver via EtherCAT. The servo driver executes the signal to drive the motor to move. The motor movement is converted into the displacement of the steel belt through the loading device 6, thereby loading the force onto the clamping and releasing device on the solar panel. (4) After each loading step is completed, the system determines whether the current target force is equal to the final target force. If the current target force is not equal to the final target force, (2) to (4) are repeated. In (2), "When loading for the first time, convert the analog quantity fed back by the S-type tension sensor of each loading device 6 at the current moment into a force value and take the average value, and replace the "based on the difference between the final target force and the average value of the current force" with "based on the difference between the final target force and the current target force", and replace all "60% of the maximum loading force in a single step" in (2) with "the maximum loading force in a single step". If the current target force is equal to the final target force, the current loading is ended.

[0070] Regarding the control method for a single loading device of the present invention Figure 7 The flowchart of its control algorithm is provided. Assume that at the k-th output signal sampling, the measured motor rotation angle is θ(k), the measured loading force is F(k), and the parameter to be identified is the system stiffness. If each sampling moment during the loading process is considered quasi-static, then after the kth sampling, the motor rotation angle θ(k), the loading force F(k), and the identified system stiffness are... The following relationship exists between them:

[0071]

[0072]

[0073] In the formula, e(k) represents the actual sampled loading force F(k) and the identified system stiffness at the k-th output signal sampling time. The predicted loading force value calculated from the motor rotation angle θ(k) The error between them. Assuming that the output variable F(k) and the observable variable θ(k) are observed k times at times i = 1, 2, ..., k, and the error of each observation is e(1), e(2), ..., e(k), then k linear equations can be established, which can be expressed in matrix form as follows:

[0074]

[0075]

[0076] The square of the sum of all identification errors in k iterations is defined as the evaluation function of the recursive least squares algorithm with a forgetting factor, which can be expressed as follows:

[0077]

[0078] The least squares estimate of the parameters corresponds to the parameters when the objective function J(k) reaches its minimum value after the k-th sampling. Let J be Find the first derivative and set it to 0, then

[0079]

[0080] Based on the above formula, the parameters after the kth sampling can be obtained. Least squares estimator:

[0081]

[0082] For online real-time parameter identification, it is necessary to... The estimation formula is transformed into a recursive algorithm, thus obtaining the recursive least squares algorithm. However, what we need to identify is the system stiffness at a certain time k. We only need the sampled values ​​near time k, instead of all historical sampled values. Therefore, we can use least squares with a forgetting factor to reduce the influence of historical data on the current time. Introducing the vectors P(k) and γ(k), we obtain the formula for the recursive least squares method with a forgetting factor:

[0083]

[0084] In the formula, λ is the forgetting factor, representing the weight of the data. By appropriately selecting the value of λ, the current time can be obtained. The optimal parameter identification value. Figure 10 The results show the online identification of the total stiffness of the system when loaded from 500N to 4000N on an actual solar array using the recursive least squares method.

[0085] For PID controllers, the discretized form is as follows:

[0086]

[0087] For the system consisting of loading device 6 and a single loading point of the solar array, ignoring parameters such as system rotational inertia and damping, and only considering the relationship between the motor output angle and the loading force, the actual stiffness of the system can be assumed to be... We hope

[0088]

[0089] That is, at any given time, K p and The product of these two factors is a constant, meaning the total gain of the system remains unchanged. Regardless of... How it changes, we need to correct K. p This keeps the total gain of the system constant, thus ensuring the stability of the control system's performance and preventing it from being affected by changes in system stiffness. The loading accuracy is ensured by adapting to changes in load. The recursive least squares identification method and the PID controller together constitute the control method for a single loading device of this invention, and its schematic diagram is shown below. Figure 7 As shown, the final control algorithm for a single loading device in this invention is as follows:

[0090]

[0091] In the formula, F i0 F represents the desired current applied force. i (k) represents the magnitude of the actual applied force fed back by the tension sensor, e i (k) represents the error term of the feedback control system. K P T is the proportional gain of the PID controller. I Let T be the integration time constant. D Let T be the differential time constant, T be the system sampling period, and k be the number of samples. M is a constant for the identified system stiffness.

[0092] Figure 8This is a schematic diagram of the control method for multiple loading devices according to the present invention. The control method for multiple loading devices is a deviation-coupled control algorithm based on average values, which is coupled with the control method for a single loading device. Its principle is to sum and average the deviations between the target force value and the current force value at each clamping point.

[0093]

[0094] Subsequently, the difference between the error at the current point and this average value is corrected by a PID controller and then passed to the deviation calculation of each loading point. The new error for each loading point is expressed as:

[0095]

[0096] In the formula, K′ P To correct the proportional gain of the PID controller for error, T′ I T′ is the integral time constant of the error-corrected PID controller. D This is the derivative time constant of the PID controller for error correction. Each point is corrected once using the average error of all loading points as the target value, thus achieving high-precision synchronous control. The final control algorithm formula, combining the control methods for a single loading device and those for multiple loading devices, is as follows:

[0097]

[0098] Thus, by nesting the average-value-based deviation coupling control algorithm into the control method for the single loading device, the two together constitute the control method for the multiple loading devices.

[0099] Figure 11 , 12 Figures 13 and 14 show the curves of the actual loading from 300N to 4000N on the solar array, combined with the control method for a single loading device, the control method for multiple loading devices, and the stepped loading strategy of this invention, and the maximum deviation curves between these curves and the points in which the loading occurs. The maximum single-step loading force at each stage is 200N. From the figures, it can be observed that:

[0100] (1) During the process of loading from 300N to 4000N, the deviation of the loading force between each point is less than 50N at any time.

[0101] (2) In each loading step, the maximum deviation of the loading force between each point always occurs in the rising phase. This is due to the inherent characteristics of the feedback controller (the error between the set value and the actual value is large during the rising phase, and the loading speed is fast). As the loading force at each loading point approaches the set value, the synchronization gradually improves, and the deviation of the loading force between each point gradually decreases.

[0102] (3) In each loading step, the loading force deviation between each point fluctuates continuously, but the fluctuation amplitude is small. This is because after the present invention adopts the control method of a single loading device, even if the stiffness of the solar array changes due to the internal flexible structure, the control system can adapt to this stiffness change, thereby ensuring synchronization.

[0103] The control method described in this invention overcomes the sudden stiffness change caused by the internal flexible structure of the solar array during loading, avoids the oscillation and divergence of the control system, and has high loading synchronization accuracy, which can meet the requirements of ground assembly of the solar array.

Claims

1. A multi-point synchronous loading device for solar wing ground assembly, characterized in that, The support frame and eight loading modules are composed of; The support frame comprises a horizontal frame and a side vertical frame, four corners of the horizontal frame are respectively provided with adjustable lifting feet for supporting the horizontal frame, and the side vertical frame is perpendicular to the horizontal frame and is fixedly installed on a side cross bar of the horizontal frame, a plurality of quick mounting plates are installed at different positions on the side vertical frame of the support frame and are used for mounting the loading modules; The loading device is used to realize the loading action, the spring support uses spring elastic force to realize gravity compensation for the loading module, so that the loading device has certain self-adaptability to installation errors, reduces contact stiffness and installation stress, the spring support is installed above the three-degree-of-freedom sliding table, and the three-degree-of-freedom sliding table is used for providing displacement adjustment of the horizontal plane XY axis and slight displacement adjustment in the vertical direction for the spring support.

2. The apparatus of claim 1, further characterized by, The loading device is installed on the upper part of the spring support, and comprises a servo motor, a harmonic reducer, a screw transmission module, an output shaft and an adaptive hook, the servo motor drives the screw transmission module to rotate through the harmonic reducer, the screw transmission module drives the nut to make linear feed motion through screw rotation, so as to drive the output shaft fixedly connected with the nut to make linear feed motion, and the output shaft of the screw transmission module drives the adaptive hook to move linearly forward and backward through the tension sensor to realize the loading action.

3. The apparatus of claim 1, further characterized by, The quick mounting plate comprises a hand screw nut, a strip-shaped pressing plate, a plate body, a flexible gasket, a bolt and a cam handle, a I-shaped groove is arranged on the surface of the plate body, during installation, the three-degree-of-freedom sliding table is positioned in the I-shaped groove of the plate body, then the strip-shaped pressing plate is rotated to a certain degree, so that the strip-shaped pressing plate is clamped and matched with the positioning groove on the bottom periphery of the three-degree-of-freedom sliding table, the strip-shaped pressing plate is pressed tightly by rotating the cam handle to pull the bolt axially and drive the hand screw nut, the quick fixing and assembly of the three-degree-of-freedom sliding table and the mounting plate are realized.

4. The apparatus of claim 1 or 2 or 3, further characterized by, The distribution positions of the plurality of quick mounting plates on the support frame have +Y assembly working conditions and -Y assembly working conditions, and the installation positions of the eight loading modules are mirror-distributed under the two working condition forms.

5. The apparatus of claim 1, further characterized by, A plurality of sliding feet are further installed on the bottom surface of the horizontal frame, the sliding feet are provided with rollers and are used to move the entire support frame. 6.A control method of a multi-point synchronous loading device for solar wing ground assembly, which is used for controlling the multi-point synchronous loading device according to any one of claims 1-5, and characterized in that, (1) first, the size of the final target force and the maximum loading force of each loading step are set, whether the parameters are set correctly is checked, and if the parameters are set correctly, the next step is performed; (2) When initially loading, the analog quantity fed back by the S-shaped tension sensor of each loading device at the current time is converted into a force value and an average value is taken, the difference between the final target force size and the average value of the current force value is compared with the size of 60% of the maximum loading force of a single step, if the difference is greater than 60% of the maximum loading force of a single step, the current target force is set as the current force value + 60% of the maximum loading force of a single step; if the difference is less than 60% of the maximum loading force of a single step, the current target force is set as the final target force; (3) After determining the size of the current target force of the next loading, the software of the upper computer inputs the collected current loading force size in the form of voltage into the analog quantity collection module through the S-shaped tension sensor of each loading device, the analog quantity collection module converts the voltage signal into a digital quantity signal, and the digital quantity signal is fed back to the upper computer through Ethernet through the EtherCAT bus between the analog quantity collection module and the motion controller; then, the current digital quantity size is converted into a force signal in the upper computer, the size of the force signal is compared with the set target loading force size, the speed signal of the motor is generated through the calculation of the control method of a single loading module and the control method of multiple loading modules, the speed signal of the motor is transmitted to the motion controller through Ethernet, and then transmitted to the servo driver of the corresponding servo motor through EtherCAT, the servo driver executes the signal to drive the servo motor to move, the movement of the servo motor is converted into the displacement of the steel belt through the loading device, and the loading of the force on the sun wing is realized; (4) After the completion of each loading step, the system judges whether the current target force is equal to the final target force, if the current target force is equal to the final target force, the current loading is ended.

7. The control method of claim 6, further characterized by, In the above step (4), if the current target force is not equal to the final target force, steps (2) to (4) are repeated, and "the difference between the final target force size and the average value of the current force value" in step (2) is replaced with "the difference between the final target force and the current target force", and all "60% of the maximum loading force of a single step" in step (2) is replaced with "the maximum loading force of a single step".

Images