Clutch with position sensor for open / close curtain motor and control method

Abstract

The application discloses a clutch with a position sensor for an open-close curtain motor and a control method, relates to the field of sensing analysis, and comprises a driving main body, an inside of the driving main body is internally provided with a clutch assembly, the clutch assembly comprises a bottom disc, a swing arm and a clutch gear, the swing arm is rotationally connected to the top end of the bottom disc, the clutch gear is installed on one side of the top end of the bottom disc, position sensor A and position sensor B are respectively installed on the left side and the right side of the top end of the bottom disc, the driving main body is used for controlling the motion state of the swing arm, the position sensor A and the position sensor B respectively collect in-place state data of the swing arm in real time at different detection points of the bottom disc, and an analysis unit is installed on the surface of the driving main body. Real-time control of the motion state of the swing arm is realized, the risk of over-adjustment or under-adjustment can be identified and compensated in time, the swing arm response is more sensitive, multiple instruction adjustment schemes are automatically generated and simulated evaluation is carried out.

Description

Technical Field

[0001] This invention relates to the field of sensor analysis technology, specifically to a clutch with a position sensor and a control method for a curtain opening and closing motor. Background Technology

[0002] With the advancement of technology, the concept of smart homes has gradually become popular, and consumers' demand for comfort and convenience in their home environment continues to rise. This has promoted the rapid development of smart curtains and smart curtain motors. In office buildings, shopping malls, or residences, smart curtain motors can be integrated with smart home devices to form a smart control ecosystem. Users can control the curtains through smartphones, voice assistants, and other devices, enhancing the convenience of life. As the smart home industry develops, there are many types of devices involving curtain motors and clutches on the market, leading to intensified competition. Companies need to continuously innovate to maintain their advantages.

[0003] Existing curtain opening and closing motor control systems typically rely on a fixed set of control parameters. This approach cannot adapt to varying needs under different working environments and practical application conditions. Consequently, in some cases, the clutch arm cannot effectively achieve precise position control, easily leading to over-adjustment or under-adjustment, which affects the opening and closing effect of the curtain. The clutch arm disengagement is unreliable, affecting the manual start function. Traditional control methods rely on simple logic judgments and control methods, lacking an adaptive and intelligent system structure. They cannot be optimized according to dynamic conditions such as load conditions, resulting in low control efficiency. Summary of the Invention

[0004] (a) Technical problems to be solved

[0005] In view of the above-mentioned shortcomings of the prior art, the present invention provides a clutch with a position sensor and a control method for a curtain opening and closing motor, which can effectively solve the problems of the prior art.

[0006] (II) Technical Solution

[0007] To achieve the above objectives, the present invention provides the following technical solution:

[0008] In a first aspect, the present invention discloses a clutch with a position sensor for a curtain opening and closing motor, comprising a drive body, wherein a clutch assembly is installed inside the drive body, the clutch assembly including a chassis, a swing arm, and a clutch gear, the swing arm being rotatably connected to the top of the chassis, a clutch gear being installed on one side of the top of the chassis, and position sensors A and B being respectively installed on the left and right sides of the top of the chassis, the drive body being used to control the movement state of the swing arm, and position sensors A and B respectively collecting real-time position state data of the swing arm at different detection points on the chassis, wherein an analysis unit is installed on the surface of the drive body, the analysis unit including:

[0009] The positioning determination module is used to receive positioning status data and determine whether the swing arm is within the predetermined position range;

[0010] The data reading module is used to read the operating parameters of the drive body and the collected parameters of position sensor A and position sensor B when the position judgment module determines that the swing arm is not in the predetermined position range.

[0011] The 3D modeling module is used to extract features from the collected parameters through a deep neural network and construct a 3D dynamic model including the swing arm's motion trajectory.

[0012] The trajectory marking module is used to mark the actual motion range of the swing arm according to the three-dimensional dynamic model, and generate multiple instruction adjustment schemes for the drive body in combination with the preset swing arm size;

[0013] The filtering unit is used to perform parameter compensation and simulate operation on several instruction adjustment schemes based on the operating sensitivity of the driving body, and filter to obtain the optimal instruction adjustment scheme.

[0014] The correction execution module is used to obtain the optimal instruction adjustment scheme provided by the screening unit and convert it into the operating electrical parameters of the drive body. The drive body executes the instructions to adjust the swing arm to a predetermined position range.

[0015] Furthermore, the screening unit has sub-modules deployed at its lower levels. These sub-modules include a risk identification module, a location simulation module, and a configuration module. The risk identification module interacts with the location simulation module and the configuration module via a wireless network.

[0016] The risk identification module is used to detect the operational sensitivity of the drive unit. Based on the detection results, it analyzes the accuracy of several instruction adjustment schemes, determines whether there is a risk of over-adjustment or under-adjustment, and compensates the parameters of several instruction adjustment schemes when the detection results show a risk.

[0017] The position simulation module receives multiple instruction adjustment schemes after compensation from the risk identification module. Combining the sensing ranges of position sensor A and position sensor B with the driving parameters of the driving body, it simulates and evaluates the multiple instruction adjustment schemes and selects the instruction adjustment scheme without negative feedback.

[0018] The configuration module is used to determine the final and unique instruction adjustment scheme from the instruction adjustment schemes without negative feedback submitted by the position simulation module.

[0019] Furthermore, the risk identification module monitors the response speed and operating current fluctuations of the drive unit in real time, generates sensitivity detection results, and adjusts the parameters of the instruction adjustment scheme using a proportional-integral compensation algorithm based on these results. The calculation formula is as follows:

[0020] ;

[0021] In the formula, Represents time The compensation amount of the driving parameters in the instruction adjustment scheme is adjusted at all times. Representative proportional compensation coefficient, Represents time Error value at time, Represents the integral compensation coefficient. Represents time from 0 to time The cumulative integral value of the error, Represents a point in time Error value at time.

[0022] Furthermore, when the position simulation module performs simulation operation evaluation, it sets a predetermined evaluation threshold, which includes the swing arm position deviation threshold and the drive body operating power threshold. Only when the simulation operation result meets the evaluation threshold is the command adjustment scheme determined to be a negative feedback-free scheme.

[0023] Furthermore, the configuration module pre-stores priority rules, which, based on the energy consumption parameters of the drive body and the weight value of the swing arm adjustment time, select the scheme with the lowest energy consumption and the shortest adjustment time from the instruction adjustment schemes without negative feedback as the final and unique instruction adjustment scheme.

[0024] Furthermore, position sensor A and position sensor B are respectively angle sensors or displacement sensors. Position sensor A is set at the starting point of the swing arm, and position sensor B is set at the ending point of the swing arm, for detecting the angle or displacement data of the swing arm at the starting point and ending point of the movement, respectively.

[0025] Furthermore, the operational logic of the 3D modeling module includes:

[0026] The key features extracted from the data reading module are received. The key features include the swing arm position data and angle data collected by position sensor A and position sensor B, as well as the operating parameters of the drive body. The operating parameters include the operating speed, torque output and current fluctuation value. The key features are then normalized.

[0027] The preprocessed key features are input into a recurrent neural network (RNN) according to the time series. The hidden layer structure of the RNN is used to perform time series analysis on the swing arm position data and the driving body operation parameters to extract the motion state change features of the swing arm at different time nodes, including velocity change, acceleration change and position offset trend.

[0028] Based on the memory unit of the recurrent neural network, combined with the spatial position data collected by position sensor A and position sensor B, the motion trajectory characteristics of the swing arm in three-dimensional space are analyzed. The trajectory characteristics include the starting position, ending position and curvature change of the intermediate path of the swing arm, forming a spatial motion distribution model of the swing arm.

[0029] The temporal features and the spatial features are fused through a fully connected layer of a recurrent neural network to generate a dynamic three-dimensional running model that includes the swing arm motion trajectory and the running state of the driving body.

[0030] The dynamic three-dimensional running model is optimized by a preset loss function. The loss function is based on the error between the actual position and the predicted position of the swing arm, and the weight parameters of the recurrent neural network are adjusted by gradient descent.

[0031] Furthermore, the positioning judgment module is electrically connected to position sensor A, position sensor B, and the data reading module; the 3D modeling module is interactively connected to the data reading module and the trajectory marking module via a wireless network; and the filtering unit is interactively connected to the trajectory marking module and the correction execution module via a wireless network.

[0032] Secondly, a control method for a curtain opening / closing motor clutch is provided, comprising the following steps:

[0033] Step 1: By using position sensors A and B installed on the clutch assembly, the position status data of the swing arm is collected in real time, and it is determined whether the swing arm is within the predetermined position range; if it is not within the predetermined position range, the operating parameters of the drive body and the collected parameters of position sensors A and B are read, and key features are extracted.

[0034] Step 2: Based on the key features, a three-dimensional operating model of the clutch assembly is constructed using a deep learning algorithm. The three-dimensional operating model includes the motion trajectory of the swing arm and the operating status information of the drive body.

[0035] Step 3: Extract the parameters of the three-dimensional running model, mark the position range of the swing arm, and generate multiple instruction adjustment schemes for the drive body within the marked position range based on the preset size of the swing arm;

[0036] Step 4: Based on the operational sensitivity detection results of the driving body, analyze the accuracy of the instruction adjustment scheme to determine whether there is a risk of over-adjustment or under-adjustment; if there is a risk, dynamically compensate the parameters of the instruction adjustment scheme through a compensation algorithm.

[0037] Step 5: Simulate and evaluate the multiple command adjustment schemes after compensation. Combine the sensing range of position sensor A and position sensor B with the driving parameters of the driving body to screen out the command adjustment schemes without negative feedback during operation, and determine the final unique command adjustment scheme from them.

[0038] Step 6: Convert the final instruction adjustment scheme into electrical parameters of the drive body, and drive the drive body to execute the instruction to adjust the swing arm to a predetermined position range.

[0039] Furthermore, in step 3, the three-dimensional running model adopts a recurrent neural network as the deep learning algorithm. Through temporal analysis and spatial feature extraction of the key features, a dynamic three-dimensional running model containing the swing arm motion trajectory and the running state of the driving body is constructed.

[0040] (III) Beneficial Effects

[0041] Compared with the known prior art, the technical solution provided by this invention has the following beneficial effects:

[0042] 1. By using position sensor A and position sensor AB to collect position information in real time, and combining it with deep learning algorithms to dynamically create a three-dimensional running model, the real-time control of the swing arm's motion state is achieved, breaking through the limitations of static parameters. The execution state is monitored through sensitivity detection, and dynamic parameter adjustment is performed using a proportional-integral compensation algorithm to ensure timely updates of feedback information, making the swing arm response more sensitive.

[0043] 2. Based on the swing arm dimensions, multiple instruction adjustment schemes are automatically generated. While providing risk identification, multiple instruction adjustment schemes are simulated and evaluated. The optimal scheme is automatically selected based on the environment and actual feedback, which enhances the system's adaptability. By comprehensively considering sensor information, compensation calculation and simulation evaluation, a complete control mechanism is formed to achieve efficient and stable control of the opening and closing curtain motor. This enables the system to maintain good operational stability under complex external conditions and improves the overall safety and reliability. Attached Figure Description

[0044] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the accompanying drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are merely some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without any creative effort.

[0045] Figure 1 This is a three-dimensional structural diagram of the present invention;

[0046] Figure 2This is a three-dimensional structural diagram of the clutch assembly, position sensor A, and position sensor B in this invention;

[0047] Figure 3 This is a top view schematic diagram of the clutch assembly, position sensor A, and position sensor B in this invention;

[0048] Figure 4 This is a schematic diagram of the overall framework of the present invention;

[0049] Figure 5 This is a schematic diagram of the screening unit in this invention.

[0050] The labels in the diagram represent: 1. Drive unit; 2. Clutch assembly; 21. Chassis; 22. Swing arm; 23. Clutch gear; 3. Position sensor A; 4. Position sensor B; 5. Analysis unit; 51. Position judgment module; 52. Data reading module; 53. 3D modeling module; 54. Trajectory marking module; 55. Filtering unit; 551. Risk identification module; 552. Position simulation module; 553. Configuration module; 56. Correction execution module. Detailed Implementation

[0051] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention.

[0052] The present invention will be further described below with reference to embodiments.

[0053] Example 1

[0054] The clutch with a position sensor used in the curtain opening and closing motor of this embodiment, such as... Figures 1-5As shown, the device includes a drive body 1, inside which a clutch assembly 2 is installed. The clutch assembly 2 includes a chassis 21, a swing arm 22, and a clutch gear 23. The swing arm 22 is rotatably connected to the top of the chassis 21. The clutch gear 23 is installed on one side of the top of the chassis 21. Position sensors A3 and B4 are respectively installed on the left and right sides of the top of the chassis 21. The drive body 1 is used to control the movement state of the swing arm 22, and the clutch gear 23 is used to engage the swing arm 22. Position sensors A3 and B4 collect the position status data of the swing arm 22 in real time at different detection points on the chassis 21. Position sensors A3 and B4 are angle sensors or displacement sensors, respectively. Position sensor A3 is set at the starting point of the swing arm 22, and position sensor B4 is set at the ending point of the swing arm 22, used to detect the angle or displacement data of the swing arm 22 at the starting and ending points of the movement, respectively. An analysis unit 5 is installed on the surface of the drive body 1. The analysis unit 5 includes:

[0055] The positioning determination module 51 is used to receive the positioning status data and determine whether the swing arm 22 is within the predetermined position range;

[0056] The data reading module 52 is triggered when the positioning judgment module 51 determines that the swing arm 22 is not in the predetermined position range, and reads the operating parameters of the drive body 1 and the acquisition parameters of the position sensor A3 and the position sensor B4.

[0057] The 3D modeling module 53 is used to extract features from the collected parameters through a deep neural network and construct a 3D dynamic model including the motion trajectory of the swing arm 22.

[0058] The trajectory marking module 54 is used to mark the actual motion range of the swing arm 22 according to the three-dimensional dynamic model, and generate multiple instruction adjustment schemes for the drive body 1 in combination with the preset swing arm 22 size.

[0059] The filtering unit 55 is used to perform parameter compensation and simulate operation on several instruction adjustment schemes based on the operational sensitivity of the drive body 1, and to filter and obtain the optimal instruction adjustment scheme. The filtering unit 55 has sub-modules deployed below it, including a risk identification module 551, a position simulation module 552, and a configuration module 553. The risk identification module 551 is interconnected with the position simulation module 552 and the configuration module 553 via a wireless network.

[0060] The risk identification module 551 is used to detect the operating sensitivity of the drive body 1, analyze the accuracy of several instruction adjustment schemes based on the detection results, determine whether there is a risk of over-adjustment or under-adjustment, and compensate the parameters of several instruction adjustment schemes when the detection results show a risk.

[0061] The position simulation module 552 is used to receive multiple instruction adjustment schemes after compensation by the risk identification module 551. Combining the sensing ranges of position sensors A3 and B4 and the driving parameters of the driving body 1, it performs simulated operation evaluation on the multiple instruction adjustment schemes and filters out the instruction adjustment schemes without negative feedback. When performing simulated operation evaluation, the position simulation module 552 sets a predetermined evaluation threshold, which includes the position deviation threshold of the swing arm 22 and the operating power threshold of the driving body 1. Only when the simulation operation result meets the evaluation threshold is the instruction adjustment scheme determined to be a scheme without negative feedback.

[0062] The configuration module 553 pre-stores priority rules. The priority rules are based on the energy consumption parameters of the drive body 1 and the weight value of the adjustment time of the swing arm 22. The scheme with the lowest energy consumption and the shortest adjustment time is selected from the instruction adjustment schemes without negative feedback as the final and unique instruction adjustment scheme.

[0063] Configuration module 553 is used to determine the final and unique instruction adjustment scheme from the instruction adjustment schemes without negative feedback submitted by position simulation module 552;

[0064] The correction execution module 56 is used to obtain the optimal instruction adjustment scheme provided by the filtering unit 55 and convert it into the operating electrical parameters of the drive body 1. The drive body 1 executes the instruction to adjust the swing arm 22 to a predetermined position range.

[0065] As a preferred embodiment of this example, Figure 4 As shown, the positioning judgment module 51 is electrically connected to the position sensor A3, the position sensor B4 and the data reading module 52, the 3D modeling module 53 is interactively connected to the data reading module 52 and the trajectory marking module 54 through a wireless network, and the filtering unit 55 is interactively connected to the trajectory marking module 54 and the correction execution module 56 through a wireless network.

[0066] The risk identification module 551 monitors the response speed and operating current fluctuations of the drive unit 1 in real time, generates sensitivity detection results, and adjusts the parameters of the command adjustment scheme based on the sensitivity detection results using a proportional-integral compensation algorithm. The calculation formula is as follows:

[0067] ;

[0068] In the formula, Represents time The amount of compensation for the drive parameters in the instruction adjustment scheme at all times, expressed in units of the electrical parameter adjustment value of the drive body. This represents the proportional compensation coefficient, a preset constant used to adjust the sensitivity of the proportional compensation. Represents time The error value at any given moment, i.e., the deviation between the actual operating position of the driving body and the target position, is expressed in angles or displacements. This represents the integral compensation coefficient, a preset constant used to adjust the cumulative effect of integral compensation. Represents time from 0 to time The cumulative integral value of the error is used to correct for long-term deviations. Represents a point in time The above formula can dynamically adjust the driving parameters based on the real-time error and the accumulation of historical errors to ensure that the swing arm 22 stably approaches the predetermined position range.

[0069] Compared with existing technologies, this embodiment significantly improves the system's intelligence and adaptability. By collecting real-time data on the positioning status of the swing arm 22 and constructing a three-dimensional dynamic model using a deep neural network, the system can accurately determine the motion state and position of the swing arm 22, thereby dynamically generating multiple instruction adjustment schemes. The introduction of operational sensitivity detection and risk assessment mechanisms can promptly identify and compensate for the risks of over-adjustment or under-adjustment, making instruction execution more stable and efficient. By comprehensively considering energy consumption parameters and adjustment time, the optimal instruction adjustment scheme can be selected, ensuring higher control accuracy and reliability of the equipment under different working environments, overcoming the shortcomings of static control, insufficient feedback, and insufficient sensitivity in existing technologies.

[0070] Example 2

[0071] At other levels, this embodiment also provides another optimized mechanism based on Embodiment 1 for a control method of the curtain opening / closing motor clutch, including the following steps:

[0072] Step 1: By using position sensors A3 and B4 installed on the clutch assembly 2, the position status data of the swing arm 22 is collected in real time, and it is determined whether the swing arm 22 is within the predetermined position range; if it is not within the predetermined position range, the operating parameters of the drive body 1 and the collected parameters of position sensors A3 and B4 are read, and key features are extracted.

[0073] Step 2: Based on key features, a three-dimensional operating model of the clutch assembly 2 is constructed using a deep learning algorithm. The three-dimensional operating model includes the motion trajectory of the swing arm 22 and the operating status information of the drive body.

[0074] Step 3: Extract the parameters of the three-dimensional running model, mark the position range of the swing arm 22, and generate multiple instruction adjustment schemes for the driving body 1 within the marked position range based on the preset size of the swing arm 22; In Step 3, the three-dimensional running model adopts a recurrent neural network as a deep learning algorithm, and constructs a dynamic three-dimensional running model containing the motion trajectory of the swing arm 22 and the running state of the driving body 1 through temporal analysis of key features and spatial feature extraction.

[0075] Step 4: Based on the operational sensitivity detection results of the driving entity 1, analyze the accuracy of the instruction adjustment scheme to determine whether there is a risk of over-adjustment or under-adjustment; if there is a risk, dynamically compensate the parameters of the instruction adjustment scheme through a compensation algorithm.

[0076] Step 5: Simulate and evaluate the multiple command adjustment schemes after compensation. Combine the sensing range of position sensor A3 and position sensor B4 with the driving parameters of drive body 1, screen out the command adjustment schemes without negative feedback during operation, and determine the final unique command adjustment scheme from them.

[0077] Step 6: Convert the final instruction adjustment scheme into electrical parameters of the drive body 1, and drive the drive body 1 to execute the instruction to adjust the swing arm 22 to the predetermined position range.

[0078] Example 3

[0079] This embodiment provides the following operating logic for a 3D modeling module 53:

[0080] The key features extracted from the data reading module 52 include the position data and angle data of the swing arm 22 collected by the position sensor A3 and the position sensor B4, as well as the operating parameters of the drive body 1, including the operating speed, torque output and current fluctuation value. The key features are then normalized.

[0081] The preprocessed key features are input into the recurrent neural network (RNN) according to the time series. The hidden layer structure of the RNN is used to perform time series analysis on the position data of the swing arm 22 and the running parameters of the driving body 1, and to extract the motion state change features of the swing arm 22 at different time nodes, including velocity change, acceleration change and position offset trend.

[0082] Based on the memory unit of the recurrent neural network, combined with the spatial position data collected by position sensor A3 and position sensor B4, the motion trajectory characteristics of the swing arm 22 in three-dimensional space are analyzed. The trajectory characteristics include the starting position, ending position and curvature change of the intermediate path of the swing arm 22, forming a spatial motion distribution model of the swing arm 22.

[0083] Temporal and spatial features are fused through a fully connected layer of a recurrent neural network to generate a dynamic three-dimensional running model that includes the motion trajectory of the swing arm 22 and the running state of the driving body 1. The dynamic three-dimensional running model reflects the motion law of the swing arm 22 in the time and spatial dimensions in real time and predicts the potential position offset of the swing arm 22 at future time nodes.

[0084] The dynamic three-dimensional running model is optimized by using a preset loss function. The loss function is based on the error between the actual position and the predicted position of the swing arm 22. The gradient descent method is used to adjust the weight parameters of the recurrent neural network to improve the prediction accuracy and stability of the dynamic three-dimensional running model. The time series analysis capability of the RNN is used to capture the dynamic changes of the swing arm 22 motion and combine it with spatial features to construct a three-dimensional model, providing a technical foundation for the precise control of the clutch system.

[0085] Working principle: This invention controls the clutch assembly 2 through the drive body 1. Specifically, the swing arm 22 is started by the drive body 1 and can disengage or engage with the clutch gear 23. The clutch gear 23 provides support. Position sensors A3 and B4 detect the position status of the swing arm 22. The position judgment module 51 determines whether the swing is within the predetermined position range. When it is determined that it is not within the predetermined position range, the data reading module 52 starts to read the operating parameters of the drive body 1 and the acquisition parameters of the position sensors A3 and B4, extracts key features, and constructs a three-dimensional running model based on the key features through the three-dimensional modeling module 53 using a deep learning algorithm. The trajectory marking module 54 extracts the parameters of the three-dimensional running model, marks the position range of the swing arm 22 based on the three-dimensional running model, and generates several instruction adjustment schemes of the drive body 1 within the marked position range based on the preset size of the swing arm 22.

[0086] The risk identification module 551 detects the operating sensitivity of the drive body 1. Based on the detection results, it analyzes the accuracy of the drive body 1's operating commands to determine whether there is over-adjustment or under-adjustment. If so, it compensates the parameters in the command adjustment scheme. The position simulation module 552 receives several command adjustment schemes compensated by the risk identification module 551. Based on the sensing range of position sensors A3 and B4 and the drive parameters of the drive body 1, it simulates and evaluates the operation of several command adjustment schemes to obtain the command adjustment scheme without negative feedback during operation. The configuration module 553 obtains the final unique command adjustment scheme. The correction execution module 56 obtains the command adjustment scheme pre-stored by the configuration module 553 and converts it into the electrical parameters of the drive body 1. The drive body 1 then executes the commands.

[0087] The above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions will not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A clutch with a position sensor for a curtain opening / closing motor, characterized in that, The device includes a drive unit, inside which a clutch assembly is installed. The clutch assembly includes a chassis, a swing arm, and a clutch gear. The swing arm is rotatably connected to the top of the chassis. The clutch gear is installed on one side of the top of the chassis. Position sensors A and B are respectively installed on the left and right sides of the top of the chassis. The drive unit controls the movement state of the swing arm. Position sensors A and B collect real-time positioning data of the swing arm at different detection points on the chassis. An analysis unit is installed on the surface of the drive unit. The analysis unit includes: The positioning determination module is used to receive positioning status data and determine whether the swing arm is within the predetermined position range; The data reading module is used to read the operating parameters of the drive body and the collected parameters of position sensor A and position sensor B when the position judgment module determines that the swing arm is not in the predetermined position range. The 3D modeling module is used to extract features from the collected parameters through a deep neural network and construct a 3D dynamic model including the swing arm's motion trajectory. The trajectory marking module is used to mark the actual motion range of the swing arm according to the three-dimensional dynamic model, and generate multiple instruction adjustment schemes for the drive body in combination with the preset swing arm size; The filtering unit is used to perform parameter compensation and simulate operation on several instruction adjustment schemes based on the operating sensitivity of the driving body, and filter to obtain the optimal instruction adjustment scheme. The correction execution module is used to obtain the optimal instruction adjustment scheme provided by the screening unit and convert it into the operating electrical parameters of the drive body. The drive body executes the instructions to adjust the swing arm to a predetermined position range.

2. The clutch with a position sensor for a curtain opening / closing motor according to claim 1, characterized in that, The screening unit has sub-modules deployed below it, including a risk identification module, a location simulation module, and a configuration module. The risk identification module interacts with the location simulation module and the configuration module via a wireless network. The risk identification module is used to detect the operational sensitivity of the drive unit. Based on the detection results, it analyzes the accuracy of several instruction adjustment schemes, determines whether there is a risk of over-adjustment or under-adjustment, and compensates the parameters of several instruction adjustment schemes when the detection results show a risk. The position simulation module receives multiple instruction adjustment schemes after compensation from the risk identification module. Combining the sensing ranges of position sensor A and position sensor B with the driving parameters of the driving body, it simulates and evaluates the multiple instruction adjustment schemes and selects the instruction adjustment scheme without negative feedback. The configuration module is used to determine the final and unique instruction adjustment scheme from the instruction adjustment schemes without negative feedback submitted by the position simulation module.

3. The clutch with a position sensor for a curtain opening / closing motor according to claim 2, characterized in that, The risk identification module monitors the response speed and operating current fluctuations of the drive unit in real time, generates sensitivity detection results, and adjusts the parameters of the command adjustment scheme based on the sensitivity detection results using a proportional-integral compensation algorithm. The calculation formula is as follows: ； In the formula, Represents time The compensation amount of the driving parameters in the instruction adjustment scheme is adjusted at all times. Representative proportional compensation coefficient, Represents time Error value at time, Represents the integral compensation coefficient. Represents time from 0 to time The cumulative integral value of the error, Represents a point in time Error value at time.

4. The clutch with a position sensor for a curtain opening / closing motor according to claim 2, characterized in that, When the position simulation module performs simulation operation evaluation, it sets a predetermined evaluation threshold, which includes the swing arm position deviation threshold and the drive body operating power threshold. Only when the simulation operation result meets the evaluation threshold is the command adjustment scheme determined to be a negative feedback-free scheme.

5. The clutch with a position sensor for a curtain opening / closing motor according to claim 2, characterized in that, The configuration module pre-stores priority rules. These priority rules, based on the energy consumption parameters of the drive body and the weight value of the swing arm adjustment time, select the scheme with the lowest energy consumption and shortest adjustment time from the instruction adjustment schemes without negative feedback as the final and unique instruction adjustment scheme.

6. The clutch with a position sensor for a curtain opening / closing motor according to claim 1, characterized in that, The position sensor A and position sensor B are respectively angle sensors or displacement sensors. Position sensor A is set at the starting point of the swing arm, and position sensor B is set at the ending point of the swing arm, and are used to detect the angle or displacement data of the swing arm at the starting point and ending point of the movement.

7. The clutch with a position sensor for a curtain opening / closing motor according to claim 1, characterized in that, The operating logic of the 3D modeling module includes: The key features extracted from the data reading module are received. The key features include the swing arm position data and angle data collected by position sensor A and position sensor B, as well as the operating parameters of the drive body. The operating parameters include the operating speed, torque output and current fluctuation value. The key features are then normalized. The preprocessed key features are input into a recurrent neural network (RNN) according to the time series. The hidden layer structure of the RNN is used to perform time series analysis on the swing arm position data and the driving body operation parameters to extract the motion state change features of the swing arm at different time nodes, including velocity change, acceleration change and position offset trend. Based on the memory unit of the recurrent neural network, combined with the spatial position data collected by position sensor A and position sensor B, the motion trajectory characteristics of the swing arm in three-dimensional space are analyzed. The trajectory characteristics include the starting position, ending position and curvature change of the intermediate path of the swing arm, forming a spatial motion distribution model of the swing arm. Temporal and spatial features are fused through a fully connected layer of a recurrent neural network to generate a dynamic three-dimensional running model that includes the swing arm's motion trajectory and the driving body's running state. The dynamic three-dimensional running model is optimized by a preset loss function. The loss function is based on the error between the actual position and the predicted position of the swing arm, and the weight parameters of the recurrent neural network are adjusted by gradient descent.

8. The clutch with a position sensor for a curtain opening / closing motor according to claim 1, characterized in that, The positioning judgment module is electrically connected to position sensor A, position sensor B and data reading module. The 3D modeling module is interactively connected to data reading module and trajectory marking module via wireless network. The filtering unit is interactively connected to trajectory marking module and correction execution module via wireless network.

9. A control method for a clutch in a curtain opening / closing motor, the method being based on the clutch according to any one of claims 1-8, characterized in that, Includes the following steps: Step 1: By using position sensors A and B installed on the clutch assembly, the position status data of the swing arm is collected in real time, and it is determined whether the swing arm is within the predetermined position range; if it is not within the predetermined position range, the operating parameters of the drive body and the collected parameters of position sensors A and B are read, and key features are extracted. Step 2: Based on the key features, a three-dimensional operating model of the clutch assembly is constructed using a deep learning algorithm. The three-dimensional operating model includes the motion trajectory of the swing arm and the operating status information of the drive body. Step 3: Extract the parameters of the three-dimensional running model, mark the position range of the swing arm, and generate multiple instruction adjustment schemes for the drive body within the marked position range based on the preset size of the swing arm; Step 4: Based on the operational sensitivity detection results of the driving body, analyze the accuracy of the instruction adjustment scheme to determine whether there is a risk of over-adjustment or under-adjustment; if there is a risk, dynamically compensate the parameters of the instruction adjustment scheme through a compensation algorithm. Step 5: Simulate and evaluate the multiple command adjustment schemes after compensation. Combine the sensing range of position sensor A and position sensor B with the driving parameters of the driving body to screen out the command adjustment schemes without negative feedback during operation, and determine the final unique command adjustment scheme from them. Step 6: Convert the final instruction adjustment scheme into electrical parameters of the drive body, and drive the drive body to execute the instruction to adjust the swing arm to a predetermined position range.

10. The control method according to claim 9, characterized in that, In step 3, the three-dimensional running model uses a recurrent neural network as the deep learning algorithm. Through temporal analysis and spatial feature extraction of the key features, a dynamic three-dimensional running model containing the swing arm motion trajectory and the running state of the driving body is constructed.

Images