Motor control method and device based on AI learning, intelligent door lock and medium

Abstract

The invention relates to the technical field of intelligent door locks, and discloses a motor control method and device based on AI learning, an intelligent door lock and a medium, the method is applied to the intelligent door lock, and the method comprises the steps of obtaining motor state data of the intelligent door lock; detecting the opening and closing state of the intelligent door lock; and AI learning is carried out according to the opening and closing state and the motor state data, target control parameters are obtained, and motor drive is controlled according to the target control parameters to carry out intelligent door lock opening and closing operation. The target control parameters of motor control are obtained by designing AI learning based on the motor state data and the opening and closing state of the intelligent door lock, the influence of overlarge current on the service life of the motor when the motor is locked can be avoided, the service life of the intelligent door lock can be prolonged, the control power consumption of the motor is reduced, and the control cost is reduced. And meanwhile, flexible motor control can be achieved through motor control AI learning, the overall control efficiency and control precision of the motor can be improved, precise door opening and closing of the door lock are achieved, and the efficient use requirement of the door lock is met.

Description

Technical Field

[0001] The present invention relates to the technical field of smart door locks, and in particular to a motor control method, device, smart door lock and medium based on AI learning. Background Art

[0002] At present, there are two ways for fully automatic intelligent door locks to achieve automatic unlocking. One way is to build the motor into the lock body electronics, and drive the lock tongue through the rotation of the reduction motor to achieve automatic door opening. This method can determine whether the lock is fully opened or closed by triggering the switch through the lock body and lock tongue, and will not cause misjudgment of the switch lock. However, due to the small space inside the lock body, the reduction motor that can be placed is a small motor, so it is difficult to achieve a large torque switch lock; even worse, once the door lock is installed, if there is friction resistance with the door frame, it will cause unlocking failure. In addition, another way is to use a conventional mechanical lock body, put the reduction motor in the handle behind the lock, and then drive the lock tongue through the lock core rotating shaft to achieve automatic door opening. This method has a larger handle space behind the lock, and the reduction motor can be made larger, and the success rate of unlocking is also higher. However, this method uses the motor's stall current to determine whether the lock tongue has completed the unlocking process, which is prone to misjudgment. Therefore, this method has a certain proportion of misjudgment of unlocking in actual applications, and the motor stall is used to judge, the current will exceed the rated current of the motor, causing damage to the reduction gear, and then forming irreversible damage, which also shortens the service life of the door lock.

[0003] Based on the above problems, the prior art improves the motor control method to achieve accurate door lock opening and closing, thereby extending the service life of the door lock. Specifically, most of the current intelligent door lock motor rotation control determines whether it is in place by detecting the stall current, and the stall current is relatively large, so it will generate a large amount of additional power consumption and waste electricity. In order to avoid the serious impact of excessive current on the life of the motor during stall, the control method of the motor on another part of the intelligent door lock is to directly drive the motor to rotate through the motor driver integrated circuit, and then add a limit switch at the position where the motor needs to stop, so that the gear of the motor output shaft can stop the motor after touching the limit switch. However, there is a problem of accuracy through the limit switch method, and there may be an angle deviation for each rotation.

[0004] In summary, the above motor control method has many defects and cannot achieve precise rotation control at any angle, which makes the control effect of the smart door lock performing door opening and closing operations through the motor poor, and it is difficult to achieve efficient control of the motor in the smart door lock. Summary of the invention

[0005] In view of this, the present invention provides a motor control method, device, smart door lock and medium based on AI learning to solve the defects of the motor of the existing smart door lock, such as high control power consumption and inflexible control angle, which makes the overall control efficiency of the motor low and seriously affects the precise control of the motor.

[0006] In a first aspect, the present invention provides a motor control method based on AI learning, which is applied to a smart door lock. The method includes: Get the motor status data of the smart door lock; Detect the open and closed status of the smart door lock; AI learning is performed based on the opening and closing status and motor status data to obtain the target control parameters, and the motor drive is controlled according to the target control parameters to perform the smart door lock switch operation.

[0007] The present invention focuses on the motor control of the smart door lock. By designing AI learning based on the motor status data and the opening and closing status of the smart door lock to obtain the target control parameters of the motor control, it can avoid the influence of excessive current on the motor life when the motor is blocked, and can effectively extend the service life of the smart door lock, thereby reducing the control power consumption of the motor. At the same time, the motor control AI learning can also realize flexible control of the motor, further improving the overall control efficiency and control accuracy of the motor, realizing precise door lock opening and closing, and meeting the needs of efficient use of the door lock.

[0008] In an optional implementation, obtaining motor status data of the smart door lock includes: Collect the pulse signal of the photoelectric sensor, which is installed on the motor of the smart door lock; Calculate the rotation angle of the motor based on the pulse signal; The operating current of the motor is measured by a preset measuring device, and the operating current and the rotation angle are determined as motor state data.

[0009] The present invention obtains motor status data by installing a photoelectric sensor on the motor of the smart door lock, which can ensure the accuracy of data acquisition and thus improve the subsequent precise control of the motor.

[0010] In an optional implementation, AI learning is performed according to the switch state and motor state data to obtain target control parameters, including: Detect whether the smart door lock is connected to the Internet; If the smart door lock is not connected to the Internet, obtain the historical operation data of the motor of the smart door lock; fit the historical operation data of the motor to obtain the AI ​​learning model; use the AI ​​learning model to make predictions and obtain the corresponding output parameters; after calibrating the output parameters, obtain the target rotation angle of the motor, record the target rotation angle obtained from each calibration and determine it as the target control parameter.

[0011] The present invention designs an AI learning process according to the current networking status of the smart door lock, and designs an adaptive process for the door lock itself when the smart door lock is not connected to the network, that is, by fitting the historical operation data of the motor to adaptively output the motor rotation angle corresponding to the next control of the door lock to open and close the door, it can avoid the phenomenon of equipment damage caused by using locked current for motor control, and improve the service life of the smart door lock to a certain extent; at the same time, the designed parameter correction feedback mechanism can ensure the accuracy of the rotation angle, thereby improving the precise control of the motor.

[0012] In an optional embodiment, the output parameters include a predicted rotation angle, a first current threshold, and a second current threshold; wherein the first current threshold represents a current value corresponding to a sudden change in the current slope, and the second current threshold represents a current value corresponding to a stable motor stall, and the first current threshold is less than the second current threshold; after calibrating the output parameters, the target rotation angle of the motor is obtained, including: Get the current working current and current rotation angle of the motor; If the current rotation angle reaches the predicted rotation angle, and the current operating current is less than the first current threshold, the predicted rotation angle is added to the preset angle threshold, and the addition result is determined as the target rotation angle of the motor; If the current rotation angle does not reach the predicted rotation angle, and the current operating current is greater than the second current threshold, the predicted rotation angle is subtracted from the preset angle threshold, and the subtraction result is determined as the target rotation angle of the motor; If the current rotation angle and the current operating current do not meet the above conditions, the predicted rotation angle is determined as the target rotation angle of the motor.

[0013] The present invention designs two mechanisms, namely, motor detection feedback and current threshold feedback, using the actual motor operation data obtained and the motor prediction control data obtained through AI learning. The motor detection feedback aims to determine the positional relationship between the current rotation angle of the motor and the predicted rotation angle. The current threshold feedback aims to determine the positional relationship between the current working current of the motor and the first current threshold and the second current threshold to correct the AI ​​learning data, and then obtain the target rotation angle of the motor. This can significantly improve the quality of the target rotation angle, and thereby achieve more accurate door opening and closing operations for the smart door lock.

[0014] In an optional implementation, if the smart door lock is connected to the Internet, the motor control method based on AI learning also includes: Obtain the historical rotation angle of the motor of the smart door lock and upload it to the preset database; Perform data fitting on the historical rotation angle of the motor to obtain the first motor rotation curve of the smart door lock; Obtaining a second motor rotation curve of a door lock of the same type as the smart door lock from a preset database; Using the second motor rotation curve to calibrate the first motor rotation curve to obtain a third motor rotation curve; A target rotation angle of the motor is determined according to the third motor rotation curve, and the target rotation angle is determined as a target control parameter.

[0015] When the smart door lock is connected to the network, the present invention has designed an adaptive process for the door lock and other networked lock bodies, that is, the corresponding motor rotation curve is obtained by fitting the historical rotation angle of the motor itself, and the motor rotation curve of other door locks of the same type obtained through the network is used for calibration, which can ensure the accuracy and rationality of the motor rotation curve, and then accurately obtain the target rotation angle, and use it to achieve precise control of the motor.

[0016] In an optional implementation, before determining the target rotation angle of the motor according to the third motor rotation curve, the motor control method based on AI learning further includes: determining whether a rotation curve of the third motor deviates from a preset threshold; If the rotation curve of the third motor deviates from the preset threshold, the process returns to the step of obtaining the motor status data of the smart door lock, and controls the smart door lock to perform early warning processing; If the third motor rotation curve does not deviate from the preset threshold, the step of determining the target rotation angle of the motor according to the third motor rotation curve is performed.

[0017] The present invention verifies the motor rotation curve obtained by networking the smart door lock, and accordingly designs a judgment on whether the curve deviates from a set threshold to ensure the validity of the curve, which helps to ensure the accuracy of the motor rotation curve.

[0018] In an optional implementation, before controlling the motor drive to perform the intelligent door lock switch operation according to the target control parameter, the motor control method based on AI learning further includes: Collect the actual status data of the motor; Determine the degree of deviation between actual state data and target control parameters; If the degree of deviation is not within an acceptable range, return to the step of obtaining the motor status data of the smart door lock; If the degree of deviation is within an acceptable range, the step of controlling the motor drive to perform the intelligent door lock switch operation according to the target control parameters is executed.

[0019] The present invention verifies the validity of the parameters by judging whether the degree of deviation between the actual state data of the motor and the target control parameters obtained by AI learning is within an acceptable range, which can further ensure the accuracy of the target control parameters, thereby improving the overall control accuracy of the motor and achieving precise door lock and door opening and closing.

[0020] In a second aspect, the present invention provides a motor control device based on AI learning, which is applied to a smart door lock, and the device includes: An acquisition module is used to obtain the motor status data of the smart door lock; A detection module, used to detect the open and closed status of the smart door lock; The control module is used to perform AI learning based on the opening and closing status and motor status data to obtain target control parameters, and control the motor drive to perform intelligent door lock switch operations based on the target control parameters.

[0021] The motor control device based on AI learning of the present invention performs AI learning based on the motor status data and the opening and closing status of the smart door lock to obtain the target control parameters of the motor, avoiding the phenomenon in the prior art that the current is too large when the motor is stalled and affects the life of the motor, helping to extend the service life of the smart door lock, thereby reducing the control power consumption of the motor, and at the same time can also achieve flexible control of the motor, improve the overall control efficiency and control accuracy of the motor, further realize accurate door lock opening and closing, and meet the needs of efficient use of the door lock.

[0022] In a third aspect, the present invention provides a smart door lock, comprising: a memory and a processor, the memory and the processor are communicatively connected to each other, computer instructions are stored in the memory, and the processor executes a motor control method based on AI learning according to the first aspect or any corresponding embodiment thereof by executing the computer instructions.

[0023] In a fourth aspect, the present invention provides a computer-readable storage medium having computer instructions stored thereon, the computer instructions being used to enable a computer to execute a motor control method based on AI learning according to the first aspect or any corresponding embodiment thereof. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] In order to more clearly illustrate the specific implementation methods of the present invention or the technical solutions in the prior art, the drawings required for use in the specific implementation methods or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are some implementation methods of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0025] Figure 1 is a flow chart of a motor control method based on AI learning according to an embodiment of the present invention; Figure 2 is a flow chart of another motor control method based on AI learning according to an embodiment of the present invention; Figure 3 It is a schematic diagram of the structure of a photoelectric sensor; Figure 4 This is a schematic diagram of the current corresponding to the motor switch door; Figure 5 It is a schematic diagram of the motor rotation curve; Figure 6 This is a schematic diagram of the structure of the motor AI learning control system; Figure 7 This is the circuit diagram of the smart door lock; Figure 8 It is the feedback flow chart of the motor AI learning control system; Fig. 9 It is a schematic diagram of the AI ​​learning process of the self-locking body; Fig.10 It is a structural diagram of the intelligent door lock networking system; Fig.11 It is a schematic diagram of the AI ​​learning process of the fitting trend of other lock bodies connected to the network; Fig.12 is a structural block diagram of a motor control device based on AI learning according to an embodiment of the present invention; Fig.13 Schematic diagram of the structure of the smart door lock according to the embodiment of the present invention. DETAILED DESCRIPTION

[0026] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of the present invention.

[0027] An embodiment of the present invention provides an embodiment of a motor control method based on AI learning. It should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a set of computer executable instructions, and although a logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0028] In this embodiment, a motor control method based on AI learning is provided, which is applied to a smart door lock. Figure 1 is a flow chart of a motor control method based on AI learning according to an embodiment of the present invention, such as Figure 1As shown, the process includes the following steps: Step S101, obtaining the motor status data of the smart door lock.

[0029] It should be noted that the motor status data of this embodiment refers to the relevant data during the operation of the motor, and its specific content and acquisition means are not limited here. The data can be adaptively determined according to actual needs and obtained by referring to the corresponding conventional methods used by those skilled in the art. For example, the motor status data includes current, and the working current of the motor can be directly measured by an ammeter, which is only for exemplary purposes.

[0030] Step S102, detecting the open and closed status of the smart door lock.

[0031] In this embodiment, the open and closed state of the smart door lock refers to the open state and closed state of the door lock, and the open and closed state of the door lock can be further identified by detecting the state of the inclined bolt. Specifically, it is determined by detecting whether the inclined bolt installed inside the lock body pops out, that is, when the inclined bolt does not pop out, the door lock is in the open state, also called the door open state; when the inclined bolt pops out, the door lock is in the closed state, also called the door closed state; wherein, whether the inclined bolt pops out or not will generate a corresponding trigger signal, and the open and closed state of the door lock can actually be determined according to the type of the received trigger signal.

[0032] Step S103, AI learning is performed according to the opening and closing state and the motor state data to obtain the target control parameters, and the motor drive is controlled according to the target control parameters to perform the smart door lock switch operation.

[0033] It should be noted that the AI ​​learning in this embodiment (for example, using Q-Learning, Deep Q-Network algorithms for reinforcement learning, multimodal learning and federated learning, etc., and the specific content of AI learning can be adaptively adjusted according to actual needs) essentially refers to the open and closed states of the door lock. By continuously learning and analyzing the door opening status and angle of each user, and combining the movement trends of other lock bodies, the target control parameters for the next time the motor controls the door lock to perform a switch operation are comprehensively obtained, and they are continuously adjusted and optimized to predict the future movement trend of the door lock.

[0034] The motor control method based on AI learning in the embodiment of the present invention designs AI learning based on motor status data and the opening and closing status of the smart door lock to obtain the target control parameters of the motor control, which can avoid the influence of excessive current on the motor life when the motor is stalled, and can effectively extend the service life of the smart door lock, thereby reducing the control power consumption of the motor. At the same time, the motor control AI learning can also realize flexible control of the motor, further improve the overall control efficiency and control accuracy of the motor, realize accurate door lock opening and closing, and meet the efficient use requirements of the door lock.

[0035] In this embodiment, a motor control method based on AI learning is provided, which is applied to a smart door lock. Figure 2 is a flow chart of another motor control method based on AI learning according to an embodiment of the present invention, such as Figure 2 As shown, the process includes the following steps: Step S201, obtaining the motor status data of the smart door lock.

[0036] Specifically, the above step S201 includes: Step S2011, collect the pulse signal of the photoelectric sensor, the photoelectric sensor is installed on the motor of the smart door lock.

[0037] It should be noted that photoelectric sensors collect data based on the photoelectric effect. When an object passes through the detection area of ​​the photoelectric sensor, the generated light signal will be captured by the sensor and an output corresponding pulse signal will be generated. The corresponding motor data can be obtained by analyzing the signal.

[0038] Step S2012, calculating the rotation angle of the motor based on the pulse signal.

[0039] In this embodiment, the rotation speed of the motor is calculated by analyzing the frequency and number of the pulse signal, and then the rotation angle of the motor is calculated according to the rotation speed.

[0040] Step S2013: Use a preset measuring device to measure the working current of the motor, and determine the working current and the rotation angle as motor state data.

[0041] In this embodiment, the specific type of the preset measuring device is not limited here and is adaptively determined according to actual needs, such as a current sensor, which is only used as an example.

[0042] In practical applications, the motor status detection often uses Hall effect sensors, which can measure the strength, direction or change of the magnetic field, and can calculate the motor's speed and angle by measuring the change of the magnetic field on the motor shaft; however, since it is more sensitive to interference from the external magnetic field, the motor status data collected by this method has low accuracy. In this embodiment, the motor status data is collected by using photoelectric sensors, which is not only convenient to design, but also has high resolution and sensitivity, and can ensure the accuracy of data acquisition. Therefore, in this embodiment, photoelectric sensors are used to collect motor status data. Specifically, the speed of the object is calculated by measuring the time difference when the object passes through the detection area. Every time the gear rotates one grid, it is blocked by light once, and a switch pulse signal is generated. The rotation speed and rotation angle of the motor are calculated by the frequency and number of switch pulses, and each rotation angle can be accurately calculated and fed back in time. Figure 3 It is a schematic diagram of the structure of the photoelectric sensor. Figure 3It can be seen that the sensor includes: a transmitter, a receiver and a detection circuit. The corresponding functions of the specific components of the sensor can be found in the calculation content of the rotation angle above, and will not be repeated here.

[0043] In the embodiment of the present invention, the motor status data is obtained by installing a photoelectric sensor on the motor of the smart door lock, which can ensure the accuracy of data acquisition and thereby improve the subsequent precise control of the motor.

[0044] Step S202: Detect the open / closed state of the smart door lock. Figure 1 Step S102 of the illustrated embodiment will not be described in detail here.

[0045] Step S203, AI learning is performed according to the opening and closing state and the motor state data to obtain the target control parameters, and the motor drive is controlled according to the target control parameters to perform the smart door lock switch operation.

[0046] Specifically, in the above step S203, AI learning is performed according to the switch state and motor state data to obtain the target control parameters, including: Step S2031, detect whether the smart door lock is connected to the network.

[0047] In this embodiment, the detection method of whether the smart door lock is connected to the network is not limited here, and can be adaptively set according to actual needs. For example, the network status of the door lock can be obtained through the status indicator light of the door lock itself, that is, in actual applications, the smart door lock will be equipped with a status indicator light, and lights of different colors or flashing frequencies can indicate the network status, such as green may indicate successful networking, and red flashing may indicate a problem with the network connection; or, by checking the router or gateway of the home where the door lock is located, that is, checking the list of connected devices, if the device name or physical address of the smart door lock can be found in the list, it means that the door lock has established a connection with the network, otherwise, it means that the door lock is not connected to the network, which is only an exemplary description.

[0048] Step S2032, if the smart door lock is not connected to the Internet, obtain the historical operation data of the motor of the smart door lock; fit the historical operation data of the motor to obtain an AI learning model; use the AI ​​learning model to make predictions and obtain corresponding output parameters; after calibrating the output parameters, obtain the target rotation angle of the motor, record the target rotation angle obtained from each calibration and determine it as the target control parameter.

[0049] In this embodiment, the output parameters include a predicted rotation angle, a first current threshold and a second current threshold; wherein the first current threshold represents the current value corresponding to a sudden change in the current slope, and the second current threshold represents the current value corresponding to the motor after it becomes stalled and stabilized, and the first current threshold is smaller than the second current threshold.

[0050] It should be noted that the first current threshold in the present embodiment is the value corresponding to the current slope mutation point; the second current threshold is the value corresponding to the stall current balance point; the two current thresholds are set to better judge the feedback motor generated angle. For example, the motor rotation angle generated by AI learning fitting is 89 degrees, but when the motor controls the door body to turn to 89 degrees, the corresponding current is still very small, that is, less than the first current threshold, indicating that the motor has not used much force and may still turn a little bit. Then let the motor turn a little more to reach the first current threshold. At this time, the actual motor rotation angle is equal to 89+0.5=89.5 degrees; in addition, if the rotation angle generated by AI learning fitting is 91 degrees, when the motor controls the door body to turn to 90 degrees, it has been detected that the current is greater than the second current threshold, which means that the current is already very large and has reached the stall current, indicating that the motor has turned to the right position, that is, it can stop rotating. At this time, the actual rotation angle of the motor is 90 degrees, that is, 90-1=90 ​​degrees. Figure 4 The schematic diagram of the corresponding current of the motor switch door. It should be noted that Figure 4 (a) and (b) correspond to the current of the motor in the door opening / closing scenario, respectively, where the first threshold is the first current threshold (which is the current mutation point before reaching the stall current, at which the current and time slopes suddenly change), and the second threshold is the second current threshold (which is the stall current point, i.e. the maximum current point when the stall current is reached). It should be explained that when the motor is about to open the door, the current must suddenly increase from a stable state; the first threshold is the starting point of the sudden increase, which is equivalent to the current mutation point. At this time, the mutation point is the point with the largest slope change in the current curve; in addition, the second threshold is the current point after the stall current is balanced. Since the actual stall current has a high and stable surge, the second threshold in this embodiment is the stable point of the stall current. For example, for a series of current data collected from a motor, they are: 0.1, 0.5, 1, 2, 1, 0.5, 0.5, 0.5, 0.6, 0.8, 0.8, 0.9, 2, 4.3, 4.1, 4.0, 4.0, 4.0; among them, 0.9 is the first current threshold (because the current value suddenly changes from 0.9 to 2); 4.0 is the second current threshold (the stall current remains stable).

[0051] In the embodiment of the present invention, an AI learning process is designed according to the current networking status of the smart door lock, and an adaptive process is designed for the door lock itself when the smart door lock is not connected to the network, that is, by fitting the historical operation data of the motor, the motor rotation angle corresponding to the next control of the door lock to open and close the door is adaptively output, which can avoid the phenomenon of equipment damage caused by using stall current for motor control, and improve the service life of the smart door lock to a certain extent; at the same time, the designed parameter correction feedback mechanism can ensure the accuracy of the rotation angle, thereby improving the precise control of the motor.

[0052] Specifically, after calibrating the output parameters in the above step S2032, the target rotation angle of the motor is obtained, including: Step A1, obtaining the current working current and current rotation angle of the motor.

[0053] In this embodiment, the specific method of obtaining the current working current and the current rotation angle of the motor is as described above and will not be repeated here.

[0054] Step A2: If the current rotation angle reaches the predicted rotation angle and the current operating current is less than the first current threshold, the predicted rotation angle is added to the preset angle threshold, and the addition result is determined as the target rotation angle of the motor.

[0055] In this embodiment, the preset angle threshold is used to adjust the data deviation obtained by AI learning, that is, when the motor does not rotate in place according to the self-learning data (i.e., the predicted rotation angle) (based on the rotation angle and the actual measured motor rotation angle, and the actual measured motor current and the two current thresholds), the predicted rotation angle is corrected by the set angle threshold (i.e., the preset angle threshold). It should be noted that the specific value of the preset angle threshold is not limited here, and it can be adaptively adjusted according to actual needs.

[0056] Step A3, if the current rotation angle does not reach the predicted rotation angle, and the current operating current is greater than the second current threshold, the predicted rotation angle is subtracted from the preset angle threshold, and the subtraction result is determined as the target rotation angle of the motor.

[0057] Step A4: if the current rotation angle and the current operating current do not satisfy the above conditions, the predicted rotation angle is determined as the target rotation angle of the motor.

[0058] It should be noted that in this embodiment, step A4 indicates that the motor has rotated in place according to the self-learning data, that is, it is in place according to the actual measured motor rotation angle, and the actual motor operation meets the two current thresholds, and no additional adjustment is required. Specifically, by using two current thresholds to check and judge the deviation of AI learning data, it is possible to avoid using the traditional method, that is, continuous stall current for interpretation, which not only improves the rotation accuracy of the motor, but also saves the duration of the stall current, further saves control power consumption, and meets the efficient control requirements of the door lock.

[0059] In the embodiment of the present invention, two mechanisms, motor detection feedback and current threshold feedback, are designed using the actual motor operation data obtained and the motor prediction control data obtained by AI learning. Among them, the motor detection feedback is intended to determine the positional relationship between the current rotation angle of the motor and the predicted rotation angle, and the current threshold feedback is intended to determine the positional relationship between the current working current of the motor and the first current threshold and the second current threshold to correct the AI ​​learning data, and then obtain the target rotation angle of the motor, which can significantly improve the quality of the target rotation angle, thereby realizing more accurate door opening and closing operations for the smart door lock.

[0060] It should be noted that the AI ​​learning of motor control in this embodiment also takes into account the data information of all networked lock bodies, and obtains the door opening and closing movement rules similar to the lock body to find out the common movement trends, and then improves the motor control of its own lock body to achieve more accurate door lock opening and closing operations. Therefore, if the smart door lock is networked, the motor control method based on AI learning in this embodiment also includes: Step B1, obtaining the historical rotation angle of the motor of the smart door lock and uploading it to a preset database.

[0061] In this embodiment, the preset database is used to provide data storage space for each networked smart door lock, and its specific content can be adaptively adjusted according to actual needs.

[0062] Step B2, performing data fitting on the historical rotation angle of the motor to obtain the first motor rotation curve of the smart door lock.

[0063] Step B3, obtaining a second motor rotation curve of a door lock of the same type as the smart door lock from a preset database.

[0064] It should be noted that, in the present embodiment, the first motor rotation curve and the second motor rotation curve are both curves composed of curve parameters representing the future trend of the motor.

[0065] Step B4: calibrate the first motor rotation curve using the second motor rotation curve to obtain a third motor rotation curve.

[0066] In a specific embodiment, Figure 5It is a schematic diagram of the motor rotation curve. As can be seen from the figure, the AI ​​learning process is suitable for the lock body of the smart door lock that has been connected to the Internet. On the one hand, it can upload its own switch lock motor rotation angle data to the cloud (this data does not involve privacy or security); on the other hand, it can obtain the motor rotation curve data of the same type of lock body as the smart door lock from the cloud, that is, obtain the motor rotation curve data of other lock bodies from the cloud: R1~Rn; collect the actual data of its own lock body, and perform AI learning training fitting based on its own data, that is, fit the future trend curve data based on its own data: R'n+1, and then AI searches for similar switch door curve data in the database; the two are combined and finally output the future trend curve data through AI learning: Rn+1, that is, predict the future rotation angle trend of this lock body Rn+1.

[0067] Step B5, determining a target rotation angle of the motor according to the third motor rotation curve, and determining the target rotation angle as a target control parameter.

[0068] In an embodiment of the present invention, when the smart door lock is connected to the network, an adaptive process is designed for the door lock and other networked lock bodies, that is, the corresponding motor rotation curve is obtained by fitting the historical rotation angle of the motor itself, and the motor rotation curve of other door locks of the same type obtained through the network is used for calibration, which can ensure the accuracy and rationality of the motor rotation curve, and then accurately obtain the target rotation angle, and use it to achieve precise control of the motor.

[0069] It should be noted that the purpose of this embodiment is to improve the effectiveness of the third motor rotation curve, and a corresponding curve calibration process is designed for this purpose. Therefore, before determining the target rotation angle of the motor according to the third motor rotation curve, the motor control method based on AI learning in this embodiment also includes: Step C1, determining whether the rotation curve of the third motor deviates from a preset threshold.

[0070] In this embodiment, the preset threshold value can be adaptively adjusted according to actual project requirements, and its specific content is not limited here.

[0071] Step C2: if the rotation curve of the third motor deviates from the preset threshold, return to the step of obtaining the motor status data of the smart door lock, and control the smart door lock to perform early warning processing.

[0072] Step C3, if the third motor rotation curve does not deviate from the preset threshold, executing the step of determining the target rotation angle of the motor according to the third motor rotation curve.

[0073] In the embodiment of the present invention, the motor rotation curve obtained by networking the smart door lock is verified, and a judgment is designed to determine whether the curve deviates from a set threshold to ensure the validity of the curve, which helps to ensure the accuracy of the motor rotation curve.

[0074] Step S2033, controlling the motor drive to perform the smart door lock switch operation according to the target control parameters.

[0075] In this embodiment, the target control parameter is the motor rotation angle corresponding to the open state or closed state of the door lock, and the motor rotation is controlled according to the rotation angle, thereby driving the door body to perform the corresponding door opening or closing operation.

[0076] The present invention verifies the validity of the parameters by judging whether the actual state data of the motor is the same as the target control parameters learned by AI, which can further ensure the accuracy of the target control parameters, thereby improving the overall control accuracy of the motor and achieving precise door lock and door opening and closing.

[0077] It should be noted that in order to ensure the accuracy of the motor rotation angle obtained by AI learning, this embodiment has designed a motor angle calibration process using measured data. Therefore, before controlling the motor drive to perform the smart door lock switch operation according to the target control parameters, the motor control method based on AI learning in this embodiment also includes: Step D1, collecting actual state data of the motor.

[0078] In this embodiment, the actual state data and the target control parameter belong to the same data type, and the specific content thereof is adaptively adjusted according to actual needs. For example, the actual state data is the motor rotation angle, also called the rotation angle.

[0079] Step D2, determining the degree of deviation between the actual state data and the target control parameter.

[0080] Step D3: If the degree of deviation is not within an acceptable range, return to the step of obtaining the motor status data of the smart door lock.

[0081] Step D4, if the degree of deviation is within an acceptable range, execute the step of controlling the motor drive to perform the intelligent door lock switch operation according to the target control parameters.

[0082] In the embodiment of the present invention, the validity of the parameters is verified by judging whether the degree of deviation between the actual state data of the motor and the target control parameters obtained by AI learning is within an acceptable range, which can further ensure the accuracy of the target control parameters, thereby improving the overall control accuracy of the motor and achieving precise door lock and door opening and closing.

[0083] In a specific embodiment, a motor AI learning control system is provided, which is integrated into a smart door lock to improve the control efficiency of the motor. Figure 6 This is a schematic diagram of the structure of the motor AI learning control system. Figure 6It can be seen that the system includes: a motor detection module, a signal trigger module, and a motor AI learning module; wherein the motor detection module is integrated on the motor, and it includes a photoelectric sensor for collecting motor operation data; the signal trigger module is arranged inside the lock body, and generates a corresponding trigger signal by popping out or pressing the oblique tongue, and judges the open or closed state of the door lock according to the trigger signal; the motor AI learning module includes an AI learning chip, which continuously learns and analyzes the door opening state and angle of each user; or the movement trend of other lock bodies, to derive the next lock body rotation angle, and continuously adjust and optimize, predict future movement trends, and issue an alert if there is a situation that exceeds the threshold.

[0084] In this embodiment, the circuit block diagram of the smart door lock integrated with the above-mentioned motor AI learning control system is as follows: Figure 7 As shown. Figure 7 It can be seen that the circuit of the smart door lock includes a motor detection adaptive learning unit and a main control identification unit; the motor detection adaptive learning unit includes a motor detection module, a signal trigger module, an AI learning module and a motor; the main control identification unit includes a main control module, a radar module, a biometric module and other modules. It should be noted that other modules include but are not limited to: Bluetooth module, WIFI module, button module, voice module, etc., which can be added according to the functional adaptability of the actual door lock.

[0085] It should be noted that, in this embodiment, the continuous adjustment and optimization of the rotation angle of the lock body is achieved through motor detection feedback and current threshold feedback. Figure 8 This is the feedback flow chart of the motor AI learning control system. Figure 8 It can be seen that the current threshold feedback is to make a comprehensive judgment based on the two set current thresholds to calibrate the motor rotation angle of the adaptive output, aiming to improve the accuracy of each lock opening and closing; in addition, the motor detection feedback is to calibrate the motor rotation angle of the adaptive output in real time according to the actual angle detected by the motor, and can accurately detect the actual rotation angle value in real time. Specifically, the design of the above two feedbacks can improve the accuracy of door opening and closing, and there is no need to wait for the stall current dwell time judgment, which greatly saves control power consumption and helps to improve the overall control efficiency and control accuracy of the motor.

[0086] It should be noted that the motor AI learning module in this embodiment is designed with two AI learning processes, namely, the AI ​​learning process of the self-lock body and the AI ​​learning process of fitting trends of other networked lock bodies. It should be noted that the AI ​​learning process of the self-lock body is for the lock of the smart door lock. The AI ​​learning fits the actual rotation angle each time the door is opened and closed, and adjusts it with the help of the current threshold. This process is essentially a data process for fitting the next unlocking for its own parameter input; the AI ​​learning process of fitting trends of other networked lock bodies is for the data integrated with other lock bodies (that is, all the data of the lock and other lock bodies in the network), observes and analyzes the door opening trajectory of other lock bodies, and sees whether a curve similar to the own lock body can be found, and the current data is used to evaluate and predict the future movement trajectory. If the fitted future trend curve deviates from the normal value, an early warning is issued (that is, assuming that the normal door opening value of the current lock body is 1, if the trend of each door opening exceeds ±20% of this value, a reminder can be issued, that is, the door opening / closing value of the lock body deviates from the corresponding normal range, then a reminder is issued, one is that the user observes whether the door lock is installed abnormally or the door body is deformed, etc., so as to perform subsequent corresponding processing.

[0087] In this embodiment, Fig. 9 It is the AI ​​learning process of the self-locking body. Fig. 9 It can be seen that the motor AI learning module records the data and results of the previous n motor detections (i.e., the rotation angle parameters and current data in the figure), performs AI learning fitting (i.e., AI learning, analysis and processing of the input data in the figure) based on the rotation angle parameters and the corresponding current data, and outputs the next motor angle: D'n+1. At the same time, it generates the first current threshold B1 (i.e., the value corresponding to the current slope mutation point) and the second current threshold B2 (i.e., the value corresponding to the stall current balance point); and makes the following judgments: ① If the rotation angle reaches D'n+1, but the current is less than B1, the calibration rotation angle is: (D'n+1) + ΔD, where ΔD is the preset angle threshold; ② If the rotation angle does not reach D'n+1, but the current is greater than B2, the calibration rotation angle is: (D'n+1)-ΔD; ③Otherwise, the rotation angle is D'n+1.

[0088] Therefore, the actual rotation angle: Dn+1=D'n+1 or (D'n+1)±ΔD), is then fed back as the input parameter of the AI ​​learning module.

[0089] It should be noted that the input data of the AI ​​learning module is the rotation angle and current parameters of the previous n times, and the training fitting outputs D'n+1, B1 and B2; then it is adjusted according to the feedback, and finally Dn+1 is output; for example, the rotation angle and current data of the switch lock are input 1 to 10 times, and the 11th pre-rotation angle D11' and the first threshold B1 and the second threshold B2 are output, and then the actual rotation angle D11 is obtained according to the feedback adjustment; then the 12th data can continue to use the relevant data of 1 to 11 times as input, output D12' and updated B1 and B2, and finally feedback adjustment is used to obtain the actual output D12.

[0090] In this embodiment, Fig.10 It is a structural diagram of a smart door lock networking system. It should be noted that each smart door lock can form a smart lock networking system by connecting to the same network, in which each smart door lock can exchange data with a server (i.e., the cloud). Fig.11 It is a schematic diagram of the AI ​​learning process of other lock bodies fitting trends in the network. Fig.11 It can be seen that the detailed implementation process of the trend fitting of other lock bodies can be understood. It should be explained that in the abnormal judgment of the future trend curve data in the figure, the mean value R* is the standard value corresponding to opening or closing the door. The specific acquisition method is not limited here. For example, it can be obtained through offline calibration, or by analyzing the future trend curve data of all lock bodies in the smart door lock networking system. It is only for exemplary explanation.

[0091] In summary, the motor control method based on AI learning in the embodiment of the present invention has the following advantages: 1. It saves the power consumption caused by the stall current, thereby saving the door lock energy and increasing the battery life; 2. The chip in the designed self-learning module can self-learn and continuously update the output, making each door opening or closing operation more accurate; 3. The designed motor detection module adopts photoelectric sensor, which is convenient to design and has high resolution, sensitivity and accuracy; 4. The designed motor detection feedback and current threshold feedback can significantly improve the accuracy of door opening and closing; 5. It can adapt to the deformation of the door body, rust of the lock body, etc., and can adaptively control the rotation angle of the motor; 6. AI can learn all the data information of the networked lock bodies, obtain the movement rules of the lock bodies, and find out the movement trends. On the one hand, it can provide an early warning reference for the next stage of rotation of the lock body. On the other hand, it can optimize the lock body. That is, AI learning and training is performed based on the data of multiple locks, which can guide the optimization of the development of the next version of the lock body.

[0092] In this embodiment, a motor control device based on AI learning is also provided, which is used to implement the above-mentioned embodiments and preferred implementation modes, and will not be repeated here. As the term "module" is used below, it can implement a combination of software and / or hardware of a predetermined function. Although the devices described in the following embodiments are preferably implemented in software, the implementation of hardware, or a combination of software and hardware, is also possible and conceivable.

[0093] The present invention provides a motor control device based on AI learning, which is applied to smart door locks, such as Fig.12 As shown, the device comprises: The acquisition module 1201 is used to obtain the motor status data of the smart door lock.

[0094] The detection module 1202 is used to detect the open and closed status of the smart door lock.

[0095] The control module 1203 is used to perform AI learning according to the opening and closing state and motor state data to obtain target control parameters, and control the motor drive to perform intelligent door lock switch operation according to the target control parameters.

[0096] In some optional embodiments, the acquisition module 1201 includes: a first acquisition submodule, a second acquisition submodule and a third acquisition submodule; wherein the first acquisition submodule is used to collect the pulse signal of the photoelectric sensor, and the photoelectric sensor is installed on the motor of the smart door lock; the second acquisition submodule is used to calculate the rotation angle of the motor based on the pulse signal; the third acquisition submodule is used to measure the working current of the motor using a preset measuring device, and determine the working current and the rotation angle as motor status data.

[0097] In some optional embodiments, the control module 1203 includes: a first control submodule, a second simulation submodule and a third control submodule; wherein the first control submodule is used to perform AI learning according to the opening and closing state and the motor state data to obtain the target control parameters, including: detecting whether the smart door lock is connected to the network; if the smart door lock is not connected to the network, obtaining the historical operation data of the motor of the smart door lock; fitting the historical operation data of the motor to obtain an AI learning model; using the AI ​​learning model to perform predictions to obtain corresponding output parameters; after calibrating the output parameters, obtaining the target rotation angle of the motor, recording the target rotation angle obtained by each calibration and determining it as the target control parameter number; a second control submodule, used to obtain the historical rotation angle of the motor of the smart door lock and upload it to a preset database if the smart door lock is connected to the network; perform data fitting on the historical rotation angle of the motor to obtain a first motor rotation curve of the smart door lock; obtain a second motor rotation curve of the same type of door lock as the smart door lock from the preset database; use the second motor rotation curve to calibrate the first motor rotation curve to obtain a third motor rotation curve; determine the target rotation angle of the motor according to the third motor rotation curve, and determine the target rotation angle as a target control parameter; a third control submodule, used to control the motor drive to perform the switch operation of the smart door lock according to the target control parameter.

[0098] In some optional embodiments, the first control submodule includes: a parameter calibration unit, which is used to calibrate the output parameters and obtain the target rotation angle of the motor, including: obtaining the current working current and the current rotation angle of the motor; if the current rotation angle reaches the predicted rotation angle, and the current working current is less than the first current threshold, then the predicted rotation angle is added to the preset angle threshold, and the addition result is determined as the target rotation angle of the motor; if the current rotation angle does not reach the predicted rotation angle, and the current working current is greater than the second current threshold, then the predicted rotation angle is subtracted from the preset angle threshold, and the subtraction result is determined as the target rotation angle of the motor; if the current rotation angle and the current working current do not meet the above conditions, the predicted rotation angle is determined as the target rotation angle of the motor.

[0099] In some optional embodiments, the second control submodule includes: a curve verification unit, which is used to determine whether the third motor rotation curve deviates from a preset threshold before determining the target rotation angle of the motor based on the third motor rotation curve; if the third motor rotation curve deviates from the preset threshold, then return to the step of obtaining the motor status data of the smart door lock, and at the same time control the smart door lock to perform early warning processing; if the third motor rotation curve does not deviate from the preset threshold, then execute the step of determining the target rotation angle of the motor based on the third motor rotation curve.

[0100] In some optional embodiments, the control module 1203 also includes: a fourth control submodule, which is used to collect actual state data of the motor before controlling the motor drive to perform the smart door lock switch operation according to the target control parameters; judging the degree of deviation between the actual state data and the target control parameters; if the degree of deviation is not within an acceptable range, returning to the step of obtaining the motor state data of the smart door lock; if the degree of deviation is within an acceptable range, executing the step of controlling the motor drive to perform the smart door lock switch operation according to the target control parameters.

[0101] The further functional description of each of the above modules is the same as that of the above corresponding embodiments and will not be repeated here.

[0102] The AI ​​learning-based motor control device in this embodiment is presented in the form of a functional unit, where the unit refers to an ASIC (Application Specific Integrated Circuit) circuit, a processor and memory that executes one or more software or fixed programs, and / or other devices that can provide the above-mentioned functions.

[0103] The motor control device based on AI learning in the embodiment of the present invention avoids the phenomenon in the prior art that the current is too large when the motor is stalled and affects the life of the motor, which helps to extend the service life of the smart door lock, thereby reducing the control power consumption of the motor. At the same time, it can also realize flexible control of the motor, improve the overall control efficiency and control accuracy of the motor, and further realize accurate door lock opening and closing, meeting the needs of efficient use of the door lock.

[0104] The present invention also provides a smart door lock. Fig.13 , Fig.13 is a schematic diagram of the structure of the above-mentioned smart door lock provided in an optional embodiment of the present invention, such as Fig.13 As shown, the smart door lock includes: one or more processors 10, a memory 20, and interfaces for connecting various components, including high-speed interfaces and low-speed interfaces. The various components are connected to each other using different buses for communication, and can be installed on a common motherboard or installed in other ways as needed. The processor can process instructions executed in the smart door lock, including instructions stored in or on the memory to display graphical information of the GUI on an external input / output device (such as a display smart door lock coupled to an interface). In some optional embodiments, if necessary, multiple processors and / or multiple buses can be used together with multiple memories and multiple memories. Similarly, multiple smart door locks can be connected, and each smart door lock provides some necessary operations (for example, determined as a server array, a group of blade servers, or a multi-processor system). Fig.13 A processor 10 is taken as an example.

[0105] The processor 10 may be a central processing unit, a network processor or a combination thereof. The processor 10 may further include a hardware chip. The hardware chip may be a dedicated integrated circuit, a programmable logic device or a combination thereof. The programmable logic device may be a complex programmable logic device, a field programmable gate array, a general purpose array logic or any combination thereof.

[0106] The memory 20 stores instructions executable by at least one processor 10, so that at least one processor 10 executes the method shown in the above embodiment.

[0107] The memory 20 may include a program storage area and a data storage area, wherein the program storage area may store an operating system, an application required for at least one function; the data storage area may store data created according to the use of the smart door lock, etc. In addition, the memory 20 may include a high-speed random access memory, and may also include a non-transient memory, such as at least one disk storage device, a flash memory device, or other non-transient solid-state storage device. In some optional embodiments, the memory 20 may optionally include a memory remotely arranged relative to the processor 10, and these remote memories may be connected to the smart door lock via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0108] The memory 20 may include a volatile memory, such as a random access memory; the memory may also include a non-volatile memory, such as a flash memory, a hard disk or a solid state drive; the memory 20 may also include a combination of the above types of memory.

[0109] The smart door lock also includes a communication interface 30, which is used for the main control chip to communicate with other smart door locks or a communication network.

[0110] A computer-readable storage medium is also provided in an embodiment of the present invention. The method according to the embodiment of the present invention can be implemented in hardware, firmware, or can be implemented as a computer code that can be recorded in a storage medium, or can be implemented as a computer code that is originally stored in a remote storage medium or a non-temporary machine-readable storage medium and will be stored in a local storage medium and downloaded through a network, so that the method described herein can be stored in such software processing on a storage medium using a general-purpose computer, a dedicated processor, or programmable or dedicated hardware. Among them, the storage medium can be a magnetic disk, an optical disk, a read-only storage memory, a random access memory, a flash memory, a hard disk or a solid-state hard disk, etc.; further, the storage medium can also include a combination of the above types of memories. It can be understood that a computer, a processor, a microprocessor main control chip or programmable hardware includes a storage component that can store or receive software or computer code. When the software or computer code is accessed and executed by a computer, a processor or hardware, the method shown in the above embodiment is implemented.

[0111] Although the embodiments of the present invention have been described in conjunction with the accompanying drawings, those skilled in the art may make various modifications and variations without departing from the spirit and scope of the present invention, and such modifications and variations are all within the scope defined by the appended claims.

Claims

1. A motor control method based on AI learning, applied to smart door locks, characterized in that: The method comprises: Get the motor status data of the smart door lock; Detecting the open and closed state of the smart door lock; AI learning is performed according to the opening and closing state and the motor state data to obtain target control parameters, and the motor drive is controlled according to the target control parameters to perform intelligent door lock switch operation.

2. The motor control method based on AI learning according to claim 1, characterized in that: The step of obtaining the motor status data of the smart door lock includes: Collecting pulse signals from a photoelectric sensor, wherein the photoelectric sensor is mounted on a motor of the smart door lock; Calculating the rotation angle of the motor based on the pulse signal; A preset measuring device is used to measure the working current of the motor, and the working current and the rotation angle are determined as motor state data.

3. The motor control method based on AI learning according to claim 2, characterized in that: The AI ​​learning is performed according to the opening and closing state and the motor state data to obtain the target control parameters, including: Detect whether the smart door lock is connected to the Internet; If the smart door lock is not connected to the Internet, the historical operation data of the motor of the smart door lock is obtained; the historical operation data of the motor is fitted to obtain an AI learning model; the AI ​​learning model is used to make predictions to obtain corresponding output parameters; after calibrating the output parameters, the target rotation angle of the motor is obtained, and the target rotation angle obtained from each calibration is recorded and determined as the target control parameter.

4. The motor control method based on AI learning according to claim 3 is characterized in that: The output parameters include a predicted rotation angle, a first current threshold, and a second current threshold; wherein the first current threshold represents a current value corresponding to a sudden change in the current slope, and the second current threshold represents a current value corresponding to a stable stall of the motor, and the first current threshold is less than the second current threshold; the target rotation angle of the motor is obtained after calibrating the output parameters, including: Obtaining the current operating current and current rotation angle of the motor; If the current rotation angle reaches the predicted rotation angle, and the current operating current is less than the first current threshold, the predicted rotation angle is added to the preset angle threshold, and the addition result is determined as the target rotation angle of the motor; If the current rotation angle does not reach the predicted rotation angle, and the current operating current is greater than the second current threshold, the predicted rotation angle is subtracted from the preset angle threshold, and the subtraction result is determined as the target rotation angle of the motor; If the current rotation angle and the current operating current do not satisfy the above conditions, the predicted rotation angle is determined as the target rotation angle of the motor.

5. The motor control method based on AI learning according to claim 3, characterized in that: If the smart door lock is connected to the network, the method further includes: Obtain the historical rotation angle of the motor of the smart door lock and upload it to a preset database; Performing data fitting on the historical rotation angle of the motor to obtain a first motor rotation curve of the smart door lock; Acquire a second motor rotation curve of a door lock of the same type as the smart door lock from the preset database; Using the second motor rotation curve to calibrate the first motor rotation curve to obtain a third motor rotation curve; A target rotation angle of the motor is determined according to the third motor rotation curve, and the target rotation angle is determined as a target control parameter.

6. The motor control method based on AI learning according to claim 5, characterized in that: Before determining the target rotation angle of the motor according to the third motor rotation curve, the method further includes: determining whether a rotation curve of the third motor deviates from a preset threshold; If the rotation curve of the third motor deviates from the preset threshold, the process returns to the step of obtaining the motor status data of the smart door lock, and controls the smart door lock to perform early warning processing; If the third motor rotation curve does not deviate from the preset threshold, the step of determining the target rotation angle of the motor according to the third motor rotation curve is performed.

7. The motor control method based on AI learning according to claim 1, characterized in that: Before controlling the motor drive to perform the intelligent door lock switch operation according to the target control parameter, the method further includes: Collect the actual status data of the motor; Determining the degree of deviation between the actual state data and the target control parameter; If the degree of deviation is not within an acceptable range, return to the step of obtaining the motor status data of the smart door lock; If the degree of deviation is within an acceptable range, the step of controlling the motor drive to perform the intelligent door lock switch operation according to the target control parameter is executed.

8. A motor control device based on AI learning, applied to smart door locks, characterized in that: The device comprises: An acquisition module is used to obtain the motor status data of the smart door lock; A detection module, used to detect the open and closed state of the smart door lock; The control module is used to perform AI learning according to the opening and closing state and the motor state data to obtain target control parameters, and control the motor drive to perform intelligent door lock switch operation according to the target control parameters.

9. A smart door lock, characterized in that: The smart door lock includes: a memory and a processor, the memory and the processor are communicatively connected to each other, the memory stores computer instructions, and the processor executes the AI ​​learning-based motor control method described in any one of claims 1 to 7 by executing the computer instructions.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a computer to execute the motor control method based on AI learning according to any one of claims 1 to 7.

Images