Intelligent control system of motor and piston air compressor driven by motor

Abstract

The application discloses an intelligent control system of a motor and a piston air compressor driven by the motor, and belongs to the technical field of motor control. The control system of the motor comprises a motor driver which provides an alternating current to make a stator of the motor generate a rotating magnetic field, and the motor driver comprises a first neural network which generates a torque correction signal Ef according to a position instruction value sequence of a rotor of the motor and a phase measurement value sequence of the motor. The application can improve the accuracy of motor control even in the case that the load of the motor continuously changes.

Description

Technical Field

[0001] This invention relates to the field of motor technology, and in particular to an intelligent control system for an electric motor and a piston air compressor driven by an electric motor. Background Technology

[0002] In servo systems for motors, controller parameters need to be adjusted in real time to ensure good dynamic and static response under different load conditions. For example, Chinese invention patent application CN115291569A discloses a servo system and its online parameter adjustment method. The method includes: determining a reference parameter sequence for the loop to be adjusted in the system, the loop to be adjusted including a speed loop and / or a position loop, the reference parameter sequence for the speed loop including speed loop gain, speed loop integral time constant and torque feedforward gain, and the reference parameter sequence for the position loop including position loop gain and speed feedforward gain; gradually adjusting the parameters of the loop to be adjusted based on the reference parameter sequence, and obtaining at least one evaluation function value corresponding to the loop to be adjusted after parameter adjustment; and determining the optimal parameters of the loop to be adjusted based on at least one evaluation function value.

[0003] However, when the load on the motor is a piston air compressor, as the piston's stroke in the cylinder changes, the density of the compressed gas in the cylinder changes continuously, causing the motor load to change continuously as well. The system and method disclosed in the above-mentioned patent application do not consider the impact of the continuous change in load on the motor, so the control accuracy of the motor is low. Summary of the Invention

[0004] To overcome the shortcomings of the prior art, the purpose of this invention is to provide an intelligent control system for an electric motor and a piston air compressor driven by an electric motor, which provides high control accuracy of the motor when the load on the motor changes continuously.

[0005] To achieve the aforementioned objective, the present invention provides a control system for an electric motor, comprising a motor driver that provides alternating current to generate a rotating magnetic field in the stator of the motor. The motor driver includes an artificial intelligence module, which comprises a data preprocessing module and a first neural network. The data preprocessing module extracts the frequency sequence from the position command value sequence and provides the position command value sequence, the frequency sequence, and the motor phase measurement value sequence to the first neural network.

[0006] The first neural network includes an input layer, a function layer, and an output layer. The input layer takes into account a sequence of position command values, a frequency sequence, and a sequence of phase measurement values ​​of the motor.

[0007] The functional relationships in the function layer are as follows:

[0008]

[0009] In the formula, Ef is the torque correction signal. The gain of the position controller for the motor driver, Here, T is the gain of the motor driver speed controller, and T is the integral time constant of the speed control. This indicates the position command value at the current time t; This indicates the position command value at the previous time t-1; This indicates the position command value at time t-2 two moments ago; for Frequency components; for Frequency components; The position command value is Phase measurement value of the motor, The position command value is The phase measurement value of the motor; Indicates the coefficient to be adjusted. The loss function represents the optimization. Indicates the number of training sessions;

[0010] The output torque correction signal Ef of the output layer.

[0011] Preferably, the motor includes a housing and a rotor shaft. Annular magnets are provided at both ends of the housing, and magnetic wheels are provided at both ends of the rotor shaft. The magnetic wheels are disposed within the cavities formed by the annular magnets, and the annular magnets and magnetic wheels repel each other along the radial direction of the rotor shaft.

[0012] Preferably, the magnetic wheel includes a flywheel disposed in the center and magnets disposed on the edge of the flywheel.

[0013] Preferably, the control system further includes a control device and a detector; the motor driver further includes an inverter, which converts the DC power supplied by the power supply into AC current and supplies it to the motor coil; a resistor is provided between the power supply and the inverter, and an inductor is connected in parallel with the resistor; the first end of the resistor is connected to ground through a first capacitor, and the second end of the resistor is connected to ground through a second capacitor; the detector is used to detect the AC current flowing through the inverter and the current flowing through the resistor and provides it to the control device; the control device includes a second neural network, which determines whether the resistor and the bridge arm of the inverter have failed based on the effective value of the AC current flowing through the motor coil and the effective value of the current flowing through the resistor provided by the detector.

[0014] Preferably, the detector includes a first current sensor, a second current sensor, a third current sensor, and a fourth current sensor. The first, second, and third current sensors are used to detect the AC current flowing through the three motor coils of the inverter, respectively. The fourth current sensor is used to detect the current flowing through the resistor and provide it to the control device. The control device includes a data processing unit and a self-competitive neural network. The data processing unit calculates the first, second, third, and fourth effective values ​​of the current based on the first, second, third, and fourth instantaneous current values ​​provided by the first, second, third, and fourth current sensors, respectively. The second neural network determines whether the resistor and the inverter's bridge arm have failed based on the first, second, third, and fourth effective values ​​of the current provided by the data processing unit.

[0015] Preferably, the second neural network includes an input layer, an intermediate layer, and an output layer. The intermediate layer arranges neurons in a two-dimensional form and learns the first current effective value, the second current effective value, the third current effective value, the fourth current effective value, the resistor, the inverter bridge arm normal operation data, short circuit data, and open circuit data in advance.

[0016] Preferably, the input layer inputs a fourth effective electrical value, a first effective current value, a second effective current value, and a third effective current value. These four effective current values ​​are then clustered with neurons in the intermediate layer. If the input data clusters near neurons that indicate the inverter arm is working normally, it indicates the inverter arm is working normally. If the input data clusters near neurons that indicate the inverter arm is short-circuited, it indicates the inverter arm is short-circuited. If the input data clusters near neurons that indicate the inverter arm is open-circuited, it indicates the inverter arm is open-circuited. Similarly, if the data clusters near neurons that indicate the inverter resistor is working normally, it indicates the inverter resistor is working normally. If the input data clusters near neurons that indicate the resistor is short-circuited, it indicates the resistor is short-circuited. If the input data clusters near neurons that indicate the resistor is open-circuited, it indicates the resistor is open-circuited.

[0017] To achieve the aforementioned objective, the present invention also provides a piston air compressor, comprising a cylinder, a piston, and a crank, characterized in that the crank is driven by a motor to cause the crank to drive the piston to reciprocate within the cylinder to compress gas and store it in a gas storage container, the control system of the motor comprising a motor driver, the motor driver providing alternating current to generate a rotating magnetic field in the stator of the motor, characterized in that the motor driver comprises an artificial intelligence module, the artificial intelligence module comprising a data preprocessing module and a first neural network, the data preprocessing module extracting the frequency sequence from the position command value sequence and providing the position command value sequence, the frequency sequence, and the phase measurement value sequence of the motor to the first neural network;

[0018] The first neural network includes an input layer, a function layer, and an output layer. The input layer takes into account a sequence of position command values, a frequency sequence, and a sequence of phase measurement values ​​of the motor.

[0019] The functional relationships in the function layer are as follows:

[0020]

[0021] In the formula, Ef is the torque correction signal. The gain of the position controller for the motor driver, Here, T is the gain of the motor driver speed controller, and T is the integral time constant of the speed control. This indicates the position command value at the current time t; This indicates the position command value at the previous time t-1; This indicates the position command value at time t-2 two moments ago; for Frequency components; for Frequency components; The position command value is Phase measurement value of the motor, The position command value is The phase measurement value of the motor; Indicates the coefficient to be adjusted. The loss function represents the optimization. Indicates the number of training sessions;

[0022] The output torque correction signal Ef of the output layer.

[0023] Preferably, the control system further includes a control device, and the motor driver further includes a motor controller; an exhaust valve is provided in the cylinder, the air inlet of the exhaust valve is connected to the cylinder, and the exhaust port is connected to the air storage container. A barometer is provided at the exhaust port, and the barometer is used to measure the air pressure in the air storage container. The air pressure value measured by the barometer is provided to the control device, and the control device provides a control signal to the motor controller according to the air pressure value measured by the barometer. The motor controller generates a position command value according to the control signal.

[0024] Compared with the prior art, the present invention has the following beneficial effects:

[0025] 1. The present invention generates a torque correction signal Ef based on the motor position command value sequence and the motor phase measurement value sequence through a first neural network, so that the motor control accuracy is also high when the motor load changes continuously.

[0026] 2. This invention provides annular magnets at both ends of the motor housing and magnetic wheels at both ends of the rotor shaft. The magnetic wheels are positioned within the cavities formed by the annular magnets. The annular magnets and magnetic wheels repel each other along the radial direction of the rotor shaft, thereby reducing friction between the rotor and the bearings and saving energy.

[0027] 3. This invention uses a detector and a second neural network to automatically determine whether a fault has occurred in the resistor or each arm of the inverter. Attached Figure Description

[0028] Figure 1 This is a schematic diagram of the composition of the motor provided by the present invention;

[0029] Figure 2 This is a power supply circuit diagram for the motor driver provided by the present invention;

[0030] Figure 3 This is a circuit diagram of the motor control system provided by the present invention;

[0031] Figure 4 This is a block diagram of the piston air compressor provided by the present invention. Detailed Implementation

[0032] To make the technical problems to be solved, the technical solutions, and the beneficial effects of the present invention clearer, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the present invention and are not intended to limit the present invention.

[0033] Figure 1 This is a schematic diagram of the motor provided by the present invention, as shown below. Figure 1As shown, the motor provided by the present invention includes a cylindrical housing 1, a stator 2 disposed within the housing, and a rotor 3 disposed in a cavity formed by the stator. Motor coils are disposed on the stator 2, and the rotor 3 is fixed on a rotor shaft 4. The rotor shaft 4 overlaps with the central axis of the housing 1. The stator 2, rotor 3, and rotor shaft 4 are coaxially fixed to the inner circumferential surface of the housing 1. Magnetic wheels 5a and 5b are respectively disposed at both ends of the rotor shaft 4.

[0034] Near the two end faces of the housing, there are a first magnetic body 9a and a second magnetic body 9b in the shape of a ring. The first magnetic body 9a and the second magnetic body 9b are fixed to the inner circumferential surface of the housing 1 coaxially with the central axis of the housing 1. The magnetic wheel 5a and the magnetic wheel 5b are respectively arranged in the cavity formed by the first magnetic body 9a and the second magnetic body 9b. The magnetic bodies and the magnetic wheels repel each other along the radial direction of the rotor shaft 4.

[0035] Rolling bearings 8a and 8b are fixed at the center of both end faces of the housing. A predetermined clearance is formed between the rolling bearings and the rotor shaft 4, as described later, so that the rolling bearings do not contact the rotor shaft 4 during periods when the rotor shaft 4 floats due to magnetic force. The rolling bearings are provided to function as bearings during device startup and to limit rotor movement exceeding a predetermined amount due to interference, etc.

[0036] Magnetic wheel 5a includes a circular plate 7a and a magnet 6a fixed to the outer periphery of the circular plate, the magnet being formed into a cylindrical shape. Magnetic wheel 5b includes a circular plate 7b and a magnet 6b fixed to the outer periphery of the circular plate, the magnet being formed into a cylindrical shape. The magnetic wheel has the function of stabilizing the rotation of the rotor through its own inertial torque. To achieve this purpose, it is preferable that the mass of the magnetic wheel be as large as possible; for this purpose, the circular plate is preferably a flywheel.

[0037] In this invention, the rotor is suspended and rotates. Therefore, in addition to reducing the friction generated during rotation, since the force is repulsive, the rotating shaft can be kept in a position of magnetic equilibrium without complex control. Furthermore, since the magnetic wheel and the rotor shaft 4 are integrated, the device can be miniaturized.

[0038] In this invention, the stator includes three sets of motor coils, namely coil U, coil V and coil W, which are provided with alternating current power to form a rotating magnetic field.

[0039] Figure 2 This is a power supply circuit diagram of the motor driver provided by the present invention, such as... Figure 2As shown, the present invention provides a power supply consisting of a solar panel 100, a charging circuit 400, a battery 500, a power supply network, and a switching circuit 800. The power supply network includes a mains power supply 700 and a rectifier and filter circuit 600. The mains power supply 700 provides alternating current (AC), and the rectifier and filter circuit 600 rectifies and filters the AC power supplied by the mains power supply 700 to generate a direct current (DC) voltage. The switching circuit is used to switch between the solar panel 100, the battery 500, and the power supply network. When solar energy is sufficient, the solar panel 100 provides power to the motor driver and charges the battery through the charging circuit 400. When the supply voltage of the solar panel 100 is lower than the supply voltage of the battery 500, the battery 500 provides power to the motor driver. When the DC power output of both the solar panel 100 and the battery 500 is lower than a set value, the mains power provides DC power to the motor driver 200. In this invention, the mains power can also be replaced by an industrial power supply network.

[0040] According to one embodiment of the present invention, the switching circuit includes diode D1, diode D2, comparator OP1, opto-isolator 801, and relay J. Relay J includes a coil and two switching switches K1 and K2. The first output terminal of the solar cell array 100 is connected to the positive terminal of diode D1, and the second output terminal is connected to ground. Figure 1 (Not shown in the image) The negative terminal of diode D1 is connected to the first stationary point of switch K2, and the moving point of switch K2 provides DC power to the motor driver; the first output terminal of battery 500 is connected to the positive terminal of diode D2, and the second output terminal is connected to ground. Figure 1 (Not shown in the image) The cathode of diode D2 is connected to the first stationary point of switch K2; the non-inverting input of comparator OP1 is connected to the cathode of diode D1, and the inverting input is connected to the movable input of adjustable resistor Rw. The output is connected to power supply Vcc2 via the LED of opto-isolator 801; the two fixed terminals of adjustable resistor Rw are connected to power supply Vcc2 and ground, respectively. In this invention, the set voltage value is taken from the movable input of adjustable resistor Rw and ground. The emitter of the transistor in opto-isolator 801 is grounded, the collector is connected to power supply Vcc1 via the coil of relay J, and the base is used to receive the light signal from the LED and convert it into an electrical signal; a diode is connected in parallel across the coil. The first stationary point of the switch K1 is left floating, and the second stationary point and the moving point are connected in series in the power supply branch of the generator fuel supply control circuit. The output terminal of the rectifier filter circuit 600 is connected to the second stationary point of the switch K2. When the DC voltage provided by the solar panel and the battery is lower than the set value, the moving point and the second stationary point of the switch K1 are connected, and the moving point and the second stationary point of the switch K2 are connected.

[0041] In this invention, when there is sufficient sunlight, and the output voltage E1 of the solar cell array 100 is higher than the output voltage E2 of the battery 500, diode D1 conducts, diode D2 is cut off, the voltage at the non-inverting input of comparator OP1 is higher than the voltage at the inverting input, comparator OP1 outputs a high level, no current flows through the light-emitting diode of the opto-isolator 801, the phototransistor does not work, no current flows through the coil of the current collector J, the moving point of K1 is connected to the first stationary point, the moving point of K2 is connected to the first stationary point, and the solar cell array 100 provides power to the motor driver. Simultaneously, the solar energy... Battery pack 100 charges battery 500. When sunlight weakens, the output voltage E1 of solar cell pack 100 is lower than the output voltage E2 of battery 500. Diode D1 is cut off, diode D2 is turned on, the voltage at the non-inverting input of comparator OP1 is higher than the voltage at the inverting input, comparator OP1 outputs a high level, no current flows through the LED of the optocoupler, the phototransistor does not work, no current flows through the coil of current collector J, the moving contact of K1 is connected to the first stationary point, the moving contact of K2 is connected to the first stationary point, and the battery 500 provides power to the motor driver. When the battery voltage E2 is lower than the input voltage at the inverting input of comparator OP1, comparator OP1 outputs a low level, current flows through the LED of the optocoupler, the phototransistor works, current flows through the coil of current collector J, the moving contact of K1 is connected to the second stationary point, the moving contact of K2 is connected to the second stationary point, and the mains power provides power E3 to the motor driver.

[0042] In this invention, the battery 500 can be any rechargeable battery, such as a lithium-ion battery, a nickel-metal hydride battery, or a nickel-cadmium battery. According to one embodiment of the invention, the charging circuit 400 can be electrically connected to the solar panel 100 via a connector, and the battery 500 can be electrically connected to the diode D2 via another connector.

[0043] Figure 3 This is a circuit diagram of the motor control system provided by the present invention, such as... Figure 3As shown, the motor control system provided by this invention includes: a filter circuit, a motor driver, a control device, and a detector. The filter circuit includes a resistor R, capacitors C1 and C2, and an inductor L forming a closed circuit. The resistor R and inductor are connected in parallel and then in series between the power supply and the motor driver. Capacitor C1 is connected between the first terminal of the resistor and ground to smooth the voltage of the power supply; that is, capacitor C1 is used to stabilize the voltage output from the power supply to the inverter. Capacitor C2 is connected between the second terminal of the resistor R and ground, and capacitor C2 is used to smooth the voltage output from the power supply to the motor driver. The filter circuit provides the DC power supply voltage to the motor driver and suppresses resonance in the closed circuit by attenuating the AC component in the current flowing through the closed circuit through the resistor. Therefore, even if the resistor fails, the DC power supply voltage is still provided to the motor driver.

[0044] The motor driver includes a motor controller and an inverter. The inverter includes four MOSFETs: T1, T2, T3, T4, T5, and T6. A freewheeling diode is placed between the source and drain of each MOSFET. The motor controller provides control signals to each MOSFET in the inverter. MOSFETs can be replaced by various semiconductor switching elements such as IGBTs (Insulated Gate Bipolar Transistors).

[0045] The detector includes a first current sensor, a second current sensor, a third current sensor, and a fourth current sensor. The first, second, and third current sensors are used to detect the AC current flowing through the three motor coils of the inverter, respectively. The fourth current sensor is used to detect the current flowing through the resistor and provide it to the control device. The control device includes a data processing unit and a self-competitive neural network. The data processing unit calculates the first, second, third, and fourth effective values ​​of the current based on the first, second, third, and fourth instantaneous current values ​​provided by the first, second, third, and fourth current sensors, respectively. The second neural network determines whether the resistor and the inverter's bridge arm have failed based on the first, second, third, and fourth effective values ​​of the current provided by the data processing unit.

[0046] The second neural network includes an input layer, an intermediate layer, and an output layer. The intermediate layer arranges neurons in a two-dimensional form and learns the first current effective value, the second current effective value, the third current effective value, the fourth current effective value, the resistor, the inverter bridge arm normal operation data, short circuit data, and open circuit data in advance.

[0047] Specifically, taking the determination of whether resistor R has malfunctioned as an example: The input layer of the second neural network is pre-input with the first, second, third, and fourth effective current values ​​for resistor R under normal operation, short circuit, and open circuit conditions. The intermediate layer learns these effective current values ​​under the three states (normal operation, short circuit, and open circuit) into its two-dimensional neurons. During the determination, the subsequent effective current values ​​are input to the input layer and clustered with the neurons in the intermediate layer. When the input data of the input layer clusters near the neurons that have been learned for normal operation of the resistor element, it indicates that resistor R is working normally. When the input data of the input layer clusters near the neurons that have been learned for short circuit of the resistor element, it indicates that resistor R is short-circuited by inductor L. When the input data of the input layer clusters near the neurons that have been learned for open circuit of the resistor element, it indicates that resistor R is open circuit.

[0048] The input layer of the second neural network is pre-inputted with the first, second, third, and fourth effective current values ​​for the inverter's bridge arm under normal operation, short circuit, and open circuit conditions. The intermediate layer pre-learns these same values ​​into its neurons. During judgment, the subsequent effective current values ​​are clustered with the neurons in the intermediate layer. If the input data clusters near the neurons for the normally operating inverter bridge arm, it indicates that the inverter bridge arm is operating normally; if it clusters near the neurons for the short circuit, it indicates that the inverter bridge arm is short-circuited; and if it clusters near the neurons for the open circuit, it indicates that the inverter bridge arm is open.

[0049] According to one embodiment of the present invention, a piston air compressor is also provided, comprising a cylinder, a piston disposed within the cylinder, a crank disposed on the piston, the crank being driven by the aforementioned motor, such that the crank drives the piston to reciprocate within the cylinder to compress gas and store it in a gas storage container. An exhaust valve is disposed within the cylinder, with its inlet connected to the cylinder and its outlet connected to the gas storage container. A barometer is disposed at the outlet, the barometer being used to measure the gas pressure in the gas storage container. The gas pressure value measured by the barometer is provided to a control device, which then provides a power supply position command value to the motor driver based on the gas pressure value measured by the barometer.

[0050] Figure 4 This is a block diagram of the motor driver provided by the present invention, as shown below. Figure 4As shown, the motor driver includes a motor controller and an inverter. The motor controller includes a position controller, a speed controller, a speed converter, an adder, an inverter, a motor phase detector, and an artificial intelligence module. The motor phase detector is used to detect the phase of the rotor of the motor M and convert it into a phase measurement value of the motor. The speed converter uses the motor's phase measurement value. Converted into motor speed information The position controller input is the position command value provided by the control system. Phase measurement value of motor M Generate speed command value The speed controller inputs the speed command value. and motor speed information The torque command value to be corrected is generated, and the torque command value to be corrected and the correction signal generated by artificial intelligence are added in an adder to produce the torque command value. The inverter operates according to the torque command value. It generates alternating current that is output to the three sets of motor coils of the motor.

[0051] The artificial intelligence module includes a data preprocessing module and a first neural network. The data preprocessing module extracts the frequency sequence from the position command value sequence and provides the position command value sequence, frequency sequence, and motor phase measurement value sequence to the first neural network.

[0052] The first neural network generates a torque correction signal Ef based on the position command value sequence, frequency sequence, and motor phase measurement value sequence.

[0053] In this invention, the neural network includes an input layer, a function layer, and an output layer, wherein the input layer receives a sequence of instruction values, a frequency sequence, and a sequence of phase measurement values ​​of the motor.

[0054] The functional relationships in the function layer are as follows:

[0055]

[0056] In the formula, Ef is the torque correction signal. For the gain of the position controller, The gain of the speed controller, where T is the integral time constant of the speed control; This indicates the position command value at the current time t; This indicates the position command value at the previous time t-1; This indicates the position command value at time t-2 two moments ago; for Frequency components; for Frequency components; The position command value is Phase measurement value of the motor, The position command value is The phase measurement value of the motor; Indicates the coefficient to be adjusted. The loss function represents the optimization. Indicates the number of training sessions;

[0057] The output torque correction signal Ef of the output layer.

[0058] This invention generates a torque correction signal Ef based on the motor's position command value sequence and the motor's phase measurement value sequence through a first neural network, thereby achieving high motor control accuracy even when the motor's load changes continuously.

[0059] It should be noted that the terms "first," "second," etc., are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Therefore, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature. In the description of this invention, "a plurality of" means two or more, unless otherwise explicitly specified. "Several" means one or more, unless otherwise explicitly specified.

[0060] In the description of this invention, it should be noted that, unless otherwise explicitly specified and limited, the term "comprising" should be interpreted broadly, for example, including not only the technical features pointed out in this invention, but also other technical features.

[0061] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely illustrative of the principles of the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the present invention as claimed. The scope of protection of this invention is defined by the appended claims and their equivalents.

Claims

1. A control system for an electric motor, comprising a motor driver that provides alternating current to generate a rotating magnetic field in the stator of the motor, characterized in that, The motor driver includes an artificial intelligence module, which includes a data preprocessing module and a first neural network. The data preprocessing module extracts a frequency sequence from the position command value sequence and provides the position command value sequence, the frequency sequence, and the phase measurement value sequence of the motor to the first neural network. The first neural network includes an input layer, a function layer, and an output layer, wherein the input layer receives the position command value sequence, the frequency sequence, and the phase measurement value sequence of the motor; The functional relationships of the function layer are as follows: ， In the formula, For torque correction signal, The gain of the position controller of the motor driver. The gain of the speed controller of the motor driver, The integral time constant for speed control; Indicates the current time Position instruction value, Indicates the previous moment Position instruction value, Indicates the first two moments Position instruction value; for , frequency components, for Frequency components; The position command value is Phase measurement value of the motor, The position command value is The phase measurement value of the motor; Indicates the coefficient to be adjusted. The loss function represents the optimization. Indicates the number of training sessions; The output layer outputs the torque correction signal. .

2. The control system for the motor according to claim 1, characterized in that, The motor includes a housing and a rotor shaft. Annular magnets are provided at both ends of the housing, and magnetic wheels are provided at both ends of the rotor shaft. The magnetic wheels are disposed within the cavities formed by the annular magnets. Along the radial direction of the rotor shaft, the annular magnets and the magnetic wheels are magnetically repelled.

3. The control system for the motor according to claim 2, characterized in that, The magnetic wheel includes a flywheel located in the center and magnets located at the edge of the flywheel.

4. The control system for the motor according to claim 3, characterized in that, The control system further includes a control device and a detector; the motor driver further includes an inverter, which converts the DC power supplied by the power supply into AC current and supplies it to the motor coil. A resistor is provided between the power supply and the inverter, and an inductor is connected in parallel with the resistor. The first end of the resistor is connected to ground through a first capacitor, and the second end of the resistor is connected to ground through a second capacitor. The detector is used to detect the AC current output by the inverter to the motor coil and the current flowing through the resistor, and provides the detection results to the control device; the control device includes a second neural network, which determines whether the resistor and the bridge arm of the inverter are faulty based on the effective value of the AC current flowing through the motor coil and the effective value of the current flowing through the resistor obtained from the detection results of the detector.

5. The control system for the motor according to claim 4, characterized in that, The detector includes a first current sensor, a second current sensor, a third current sensor, and a fourth current sensor. The first current sensor, the second current sensor, and the third current sensor are used to detect the AC current output by the inverter to the three motor coils, respectively. The fourth current sensor is used to detect the current flowing through the resistor and provide the detection result to the control device. The control device further includes a data processing unit, wherein the second neural network is a self-competitive neural network; the data processing unit calculates the first effective value, the second effective value, the third effective value, and the fourth effective value of the current based on the first instantaneous value, the second instantaneous value, the third instantaneous value, and the fourth instantaneous value of the current provided by the first current sensor, the second current sensor, the third current sensor, and the fourth current sensor, respectively; the second neural network determines whether the resistor and the bridge arm of the inverter have failed based on the first effective value, the second effective value, the third effective value, and the fourth effective value of the current provided by the data processing unit.

6. The control system for the motor according to claim 5, characterized in that, The second neural network includes an input layer, an intermediate layer, and an output layer. The neurons in the intermediate layer are arranged in a two-dimensional form and are pre-learned to have the first, second, third, and fourth effective current values ​​corresponding to the resistor and the inverter bridge arm being in normal working state, short circuit state, and open circuit state, respectively.

7. The control system for the motor according to claim 6, characterized in that, The input layer inputs the first effective value of the current, the second effective value of the current, the third effective value of the current, and the fourth effective value of the current. The second neural network clusters the first effective value of the current, the second effective value of the current, the third effective value of the current, and the fourth effective value of the current with the neurons of the intermediate layer. When the input data of the input layer clusters near the neurons corresponding to the learned normal working state of the inverter bridge arm, it indicates that the inverter bridge arm is working normally; when the input data of the input layer clusters near the neurons corresponding to the learned short-circuit state of the inverter bridge arm, it indicates that the inverter bridge arm is short-circuited; when the input data of the input layer clusters near the neurons corresponding to the learned open-circuit state of the inverter bridge arm, it indicates that the inverter bridge arm is open-circuited. When the input data of the input layer clusters to the vicinity of the neurons corresponding to the normal working state of the learned resistor, it indicates that the resistor is working normally. When the input data of the input layer clusters near the neurons corresponding to the learned short-circuit state of the resistor, it indicates that the resistor is in a short-circuit state; when the input data of the input layer clusters near the neurons corresponding to the learned open-circuit state of the resistor, it indicates that the resistor is open-circuit.

8. A piston-type air compressor, comprising a cylinder, a piston, and a crank, characterized in that, The crank is driven by an electric motor, which causes the crank to drive the piston to reciprocate in the cylinder, thereby compressing the gas and storing it in the gas storage container. The control system of the motor includes a motor driver that provides alternating current to generate a rotating magnetic field in the stator of the motor. The motor driver includes an artificial intelligence module that includes a data preprocessing module and a first neural network. The data preprocessing module extracts a frequency sequence from a position command value sequence and provides the position command value sequence, the frequency sequence, and the phase measurement value sequence of the motor to the first neural network. The first neural network includes an input layer, a function layer, and an output layer, wherein the input layer receives the position command value sequence, the frequency sequence, and the phase measurement value sequence of the motor; The functional relationships of the function layer are as follows: ， In the formula, For torque correction signal, The gain of the position controller of the motor driver. The gain of the speed controller of the motor driver, The integral time constant for speed control; Indicates the current time Position instruction value, Indicates the previous moment Position instruction value, Indicates the first two moments Position instruction value; for , frequency components, for Frequency components; The position command value is Phase measurement value of the motor, The position command value is The phase measurement value of the motor; Indicates the coefficient to be adjusted. The loss function represents the optimization. Indicates the number of training sessions; The output layer outputs the torque correction signal. .

9. The piston air compressor according to claim 8, characterized in that, The control system of the motor also includes a control device, and the motor driver also includes a motor controller; the cylinder is provided with an exhaust valve, the air inlet of the exhaust valve is connected to the cylinder, the exhaust outlet of the exhaust valve is connected to the air storage container, and a barometer is provided at the exhaust outlet; The barometer is used to measure the air pressure in the air storage container and provide the measured air pressure value to the control device. The control device provides a control signal to the motor controller based on the air pressure value, and the motor controller generates a position command value based on the control signal.

Images