CN116780956B - A control method for self-learning DC brushless motor Hall position based on vector algorithm - Google Patents

Abstract

The invention discloses a control method for self-learning the Hall position of a DC brushless motor based on a vector algorithm, which comprises the following steps: determining the relation between the power-on sequence of the direct-current brushless motor and the Hall sensor to obtain a Hall sector, configuring parameters of an upper computer, and then sending an instruction to a control board; after receiving the instruction, the control board firstly learns the Hall sector by itself, and stores the Hall sector obtained by self-learning into FLASH in the chip; then, after the self-learning of the Hall sector is successful, the self-learning of the FOC algorithm is carried out, the electric speed of the motor operation is calculated through the sector, and the current angle is calculated through the electric speed; in the FOC algorithm, fixed parameters Ud and Uq are set, three paths of PWM waves are generated through the calculated angle values and are input into a three-phase stator of the motor, and the rotor is driven to rotate; through self-learning of various parameters of the direct current brushless motor in the early stage, the learned parameters are valued, so that the bus current value is collected when the motor rotates, whether the self-learning is successful or not is judged, and the rotating performance of the motor is optimized by adopting a vector control FOC algorithm; and secondly, the learned motor parameters are stored into the chip to obtain a FLASH region, so that the function of power-off protection is achieved.

Description

A control method for self-learning DC brushless motor Hall position based on vector algorithm

Technical field

The invention belongs to the technical field of brushless DC motors, and specifically relates to a control method for self-learning the Hall position of a brushless DC motor based on a vector algorithm.

Background technique

Various scenes in life are inseparable from the participation of motors. In the field of control, if there is a Hall sensor at the back end of the motor, if you want to make the brushless DC motor rotate, you need to know the three-phase windings of the brushless DC motor ( The position of the Hall sensor (HALLA, HALLB, HALLC) corresponding to U phase, V phase, W phase), and how to better control the rotation of the motor and improve the working efficiency of the motor are issues that must be considered. Usually , the traditional control methods include the following:

1. View the back electromotive force waveform of the three-phase voltage corresponding to the signal fed back by the Hall sensor through an oscilloscope to determine their one-to-one correspondence. Finally, write the code to control the rotation of the brushless DC motor. In addition, for micro For the traditional Brushless Direct Current Motor (BLDC), the traditional method uses six-step commutation control. The angle between the direction of the torque generated by the stator winding and the position of the rotor is within [60°, 120° °], causing the final motor output torque to fluctuate, which is not the optimal control strategy.

2. Under the control method with Hall sensor, facing a brand-new brushless DC motor, it is necessary to debug to obtain the position of Hall and the zero position of the motor. Through continuous debugging, find the position of the motor corresponding to the Hall sector. Angle value, thereby using the FOC algorithm to control the motor rotation. However, its complicated debugging process and the designed system can only be applied to the same type of motor and are not universal.

Contents of the invention

The purpose of this invention is to design a control method for self-learning the Hall position of a brushless DC motor based on a vector algorithm. By combining the host computer and the control board, the motor can self-learn the position of the Hall sensor and store it in the FALSH of the control board. area, data will not be lost when the control board is powered off.

In order to achieve the above objects, the technical solutions adopted by the present invention are as follows:

S1: After the host computer configures the parameters, it sends instructions to the control board

S2: After receiving the instruction, the control board first performs self-learning of the Hall sectors and stores the Hall sectors obtained by self-learning into the FLASH in the chip;

S3: Then after the self-learning of the Hall sector is successful, the FOC algorithm is self-learning, the electrical speed of the motor is calculated through the sector, and the current angle value is calculated through the electrical speed.

S4: In the FOC algorithm, fixed parameters Ud and Uq are set, and through the calculated angle values, three PWM waves are generated and input into the three-phase stator of the motor to drive the rotor to rotate.

Through the above self-learning algorithm, any type of motor can be successfully rotated under the vector control FOC algorithm.

The invention has the following beneficial effects:

The system designed by this invention uses self-learning of various parameters of the brushless DC motor in the early stage, and takes the values of the learned parameters to allow the motor to collect the bus current value when rotating to determine whether the self-learning is successful and adopts the vector control FOC algorithm. Optimize the performance of motor rotation; secondly, store the learned motor parameters inside the chip to obtain the FLASH area, which plays the role of power-off protection; finally, through the combination of the upper computer and the lower computer, the operation of the UI interface can be For a more simplified experience, the real-time changes in parameters of the brushless DC motor during the self-learning process can be detected in real time through the host computer.

Through reasonable design and intelligent control strategy, the present invention can better complete the self-learning function of the DC brushless motor, save the cost of separate debugging for each DC brushless motor, and the designed algorithm has better Adaptability and robustness to meet the needs of different products.

Any type of brushless DC motor does not need to go through preliminary testing, debugging, coding, etc. The designed algorithm allows the motor to self-learn its parameters.

Description of the drawings

Figure 1 is a diagram showing the relationship between the installation position of the Hall sensor of the brushless DC motor and the position of the motor rotor.

Figure 2 is a diagram of the relationship between the Hall sequence and the final Hall's sector Hall state value HALLState.

Figure 3 is a diagram showing the relationship between back electromotive force and HALLState.

Figure 4 is the flow chart of the brushless DC motor self-learning algorithm of the Hall sensor.

Figure 5 is a flow chart of the sector sequence obtained in the second self-learning.

Figure 6 is the relationship between the final θ _result and the Hall sector.

Figure 7 is the interface diagram of the host computer.

Detailed ways

The present invention will be further described below in conjunction with the accompanying drawings:

With the development of artificial intelligence, designed products need to be more intelligent and stronger, and the algorithm designed in the present invention can perform self-learning on any brushless DC motor with a Hall sensor. The algorithm allows the motor to learn its own various parameters, and can adaptively optimize the rotation of the motor through vector control. It uses the algorithm of Hall sector to solve the angle value, and controls the motor through the vector control FOC algorithm to make the motor During operation, greater torque is generated, current loss is reduced, and work efficiency is improved. The upper computer software is written in C# language. The interface operation can bring a more intuitive experience. The communication between the upper computer and the lower computer makes the designed system More simplicity.

With reference to Figures 1-7, the present invention discloses a control method for self-learning the Hall position of a brushless DC motor based on a vector algorithm. The method is as follows:

S1: Determine the relationship between the power-on sequence of the brushless DC motor and the Hall sensor to obtain the Hall sector, host computer configuration parameters, and then send instructions to the control board;

S2: After receiving the instruction, the control board first performs self-learning of the Hall sectors and stores the Hall sectors obtained by self-learning into the FLASH in the chip;

The specific steps of self-learning of the Hall sector: through the sequence of powering on the DC brushless motor, let the rotor move to the specified position, and obtain the Hall state HALLState of the position; follow the sequence of powering on, each time lasting 100ms, in The magnetic moment generated by the stator within 100ms causes the rotor to move, and the obtained HALLState is stored in the two-dimensional array BUFF. There are six commutations in one cycle, which takes 600ms, self-learning 10 times, and a six-step commutation control method. , detect whether the bus current value is less than the set value, and verify whether the self-learning is successful;

S3: Then after the self-learning of the Hall sector is successful, the self-learning of the FOC algorithm is performed, the electrical speed of the motor is calculated through the sector, and the current angle value is calculated through the electrical speed;

S4: In the FOC algorithm, fixed parameters Ud and Uq are set, and through the calculated angle values, three PWM waves are generated and input into the three-phase stator of the motor to drive the rotor to rotate.

For the above method, the present invention provides specific examples as follows:

The research object of this invention is that the installation positions of three Halls are 120° different, and 6 different Hall signal combinations can be output, corresponding to 6 different areas respectively. The installation position of the Hall sensor of the brushless DC motor and the position of the motor rotor are shown in Figure 1.

The relationship between the power-on sequence of the brushless DC motor and the Hall sensor is as shown in the table below, where "×" means that the phase is closed, and "√" means that the phase is on.

The three-phase Hall signal is a switching signal. The high level is recorded as 1 and the low level is 0. The Hall sector value can be calculated as follows:

HALLState _Sum =HALL _A +2*HALL _B +4*HALL _C

The Hall order of HALLState _Sum is 5-4-6-2-3-1, which can be sorted in order (0-5) according to the actual situation to facilitate angle value calculation and processing. The relationship between the Hall order and the final Hall's sector Hall state value HALLState is shown in 2. The upper curve in the figure is the sector value of HALLState, and the lower curve is the figure after sorting in order.

A's relative neutral point back electromotive force zero-crossing point (from positive to negative) is essentially the absolute zero position of the angle value. The formula in the back electromotive force formula (surface-mounted brushless DC motor) is as follows:

Ea, Eb, and Ec are the back electromotive force of the three-phase neutral points of the stator winding of the brushless DC motor. ω _e is the electrical angular velocity, ψ _f is the flux linkage of the rotor, and θ _e is the angle value. The relationship between the back electromotive force and HALLState can be obtained. The relationship is shown in Figure 3. The upper curve in the figure is the sector value of HALLState, and the lower curve is the back electromotive force of phase A. It can be obtained that the intersection point where the back electromotive force changes from positive to negative is the zero electrical angle position, and the corresponding sector The area is called the zero degree sector.

The present invention first needs to identify the Hall sector of the DC brushless motor to lay the foundation for the subsequent use of the FOC algorithm. Through the sequence of powering on the DC brushless motor, the rotor is moved to the designated position and the Hall state of the position is obtained. HALLState, according to the power-on sequence, lasts for 100ms each time. Within 100ms, the magnetic moment generated by the stator causes the rotor to move, and the obtained HALLState is stored in the two-dimensional array BUFF. There are six commutations in one cycle, which takes 600ms. Learn 10 times to improve the success rate of self-learning, and use the six-step commutation control method to detect whether the bus current value is less than the set value to verify whether the self-learning is successful. The self-learning algorithm flow chart of the DC brushless motor based on the Hall sensor of the present invention is shown in Figure 4.

After obtaining the Hall position, use the FOC vector algorithm to rotate the motor. First, you need to obtain the sector corresponding to the zero-angle position of the motor. The first self-learning algorithm obtains the one-to-one Hall wiring harness corresponding to the three phases of the motor stator winding. The second self-learning algorithm obtains the position of the zero-degree sector and knows the sequence of sector changes corresponding to the forward and reverse rotation of the motor.

For the FOC vector algorithm, the present invention obtains the displacement angle of this sector by solving the electrical angular velocity of the previous sector and integrating the electrical angular velocity of the previous sector. The flow chart of the sector sequence obtained in the second self-learning is shown in Figure 5. Before the motor starts, the motor may be in a stationary state. We need to read the three-phase Hall switch signal first to determine the Hall sector angle where the rotor is currently located. The maximum deviation between this angle and the actual angle is ±30°. After the motor rotates stably, 6 accurate Hall sector angles can be obtained within one electrical cycle. When calculating the speed, you can use the signal capture function to trigger an interrupt, and read the counting results in the interrupt for speed calculation, clear the count, and enter the next sector count. When the motor rotates for one electrical cycle, Hall produces 6 sector changes. Let ω _e represent the electrical angular speed of the motor, and the calculation formula is:

In the formula, RegisterCnt is the read timer count value, and f _timer is the frequency of the timer configuration. The mechanical speed is (1/P) times the electrical angular speed, and the formula is:

In the formula, P is the number of pole pairs of the motor, and ω _m represents the mechanical angular speed of the motor.

In the present invention, the solution of the Hall angle is achieved by integrating the velocity. First, use a timer to calculate the time it takes to walk through a Hall sector, T _oneSec (T method speed measurement). Secondly, use T _oneSec to calculate the angle value Δθ _e that needs to be traveled within one carrier cycle (T _carrier ). Finally, within each carrier interruption, the angle value is accumulated each time to achieve continuous integration of the angle. The Δθ _e here is estimated from the speed of the previous sector. Every time the Hall sector jumps, the angular velocity ω _e will be updated and the register value will be cleared. In order to facilitate the self-learning process of the brushless DC motor, a fixed deviation angle is used as θ _error for the deviation angle of the Hall sector. Then the real angle value finally calculated through the Hall sector is:

θ _result =θ _Cumulate +θ _error

In the formula, θ _Cumulate is the angle value obtained by integrating the Hall sector, θ _error is the fixed deviation value of 30°, and θ _result is the final angle value. Check the relationship between the final θ _result and the Hall sector through the J-Scope software. Figure 6 shows a fixed angle value in the initial stage. The magnetic moment generated by the stator drags the rotor to rotate. After dragging, the chip The register in can capture the transition of the Hall sector and start recording the time spent in each sector. After that, the electrical speed can be solved and the speed can be integrated to calculate the current angle value.

Input the final θ _result into the SVPWM (Space Vector Pulse Width Modulation) code. By fixing Ud equal to 0 and Uq equal to 10000, a three-phase PWM wave that changes with time can be generated, and finally a rotating vector is synthesized to drive the motor. The rotor spins.

The present invention adopts more convenient interface operations, allowing users to better understand the functions of the product. The UI interface is designed as a host computer. By defining different buttons, the host computer can communicate with the control panel by clicking the button. , the control board executes the corresponding action after getting the corresponding instruction. The interface of the host computer is shown in Figure 7.

The entire method does not require preliminary testing, debugging, coding, etc. for any type of brushless DC motor. The designed algorithm allows the motor to self-learn its parameters. If the device is powered off, since the parameter information is stored in the FLASH area of the chip in advance, the data will not be lost along with the power outage, and the experience gained from self-learning will be stored. The running status of the motor is detected in real time and displayed on the host computer in real time, so that the user can feel the parameter changes of the motor more intuitively. Through multiple self-study training sessions, the success rate of self-study is increased.

The above are all preferred embodiments of the present invention. For those of ordinary skill in the art, without departing from the principles of the present invention, modifications to various equivalent forms of the present invention all fall within the scope of the appended claims of this application. protected range.

Claims (6)

1. A control method for self-learning the Hall position of a brushless DC motor based on a vector algorithm, which is characterized in that: the method is as follows: S1: Determine the relationship between the power-on sequence of the brushless DC motor and the Hall sensor to obtain the Hall sector, host computer configuration parameters, and then send instructions to the control board; S2: After receiving the instruction, the control board first performs self-learning of the Hall sectors and stores the Hall sectors obtained by self-learning into the FLASH in the chip; The specific steps of self-learning of the Hall sector: through the sequence of powering on the DC brushless motor, let the rotor move to the specified position, and obtain the Hall state HALLState of the position; follow the sequence of powering on, each time lasting 100ms, in The magnetic moment generated by the stator within 100ms causes the rotor to move, and the obtained HALLState is stored in the two-dimensional array BUFF. There are six commutations in one cycle, which takes 600ms, self-learning 10 times, and a six-step commutation control method. , detect whether the bus current value is less than the set value, and verify whether the self-learning is successful; S3: Then after the self-learning of the Hall sector is successful, the self-learning of the FOC algorithm is performed, the electrical speed of the motor is calculated through the sector, and the current angle value is calculated through the electrical speed; S4: In the FOC algorithm, fixed parameters Ud and Uq are set, and through the calculated angle values, three PWM waves are generated and input into the three-phase stator of the motor to drive the rotor to rotate; The solution for the Hall angle is achieved by integrating the velocity: First, use a timer to calculate the time T _oneSec it takes to walk through a Hall sector, and measure the speed with the T method; Secondly, use T _oneSec to calculate the angle value Δθ _e that needs to be traveled within a carrier period T _carrier ; Finally, within each carrier interruption, the angle value is accumulated each time to achieve continuous integration of the angle; The Δθ _e here is estimated from the speed of the previous sector. Each time the Hall sector jumps, the angular velocity ω _e will be updated and the register value will be cleared; the fixed deviation angle is θ _error , and the Hall sector will eventually pass the angular velocity ω e The real angle value calculated by the sector is: θ _result =θ _Cumulate +θ _error In the formula, θ _Cumulate is the angle value obtained by integrating the Hall sector, θ _error is the fixed deviation value of 30°, and θ _result is the final angle value. 2. A control method for self-learning DC brushless motor Hall position based on vector algorithm according to claim 1, characterized in that: the installation positions of three Halls used in S1 differ by 120°, and 6 different Hall positions are output. Hall signal combinations correspond to 6 different areas. The specific relationship between the power-on sequence of the brushless DC motor and the Hall sensor is as follows: Among them, "×" means that the phase is closed, "√" means that the phase is turned on, the high-level signal is 1, and the low-level signal is 0. 3. A control method for the Hall position of a brushless DC motor based on vector algorithm self-learning according to claim 1, characterized in that: the self-learning of the FOC algorithm is divided into two times, and the first self-learning algorithm obtains the motor stator. The three-phase one-to-one Hall wiring harness of the windings can obtain the position of the zero-degree sector through self-learning for the second time, and know the sequence of sector changes corresponding to the forward and reverse rotation of the motor. 4. A control method for self-learning DC brushless motor Hall position based on vector algorithm according to claim 2 or 3, characterized in that: before the motor runs, it is necessary to read the three-phase Hall switch signal to determine the current position of the rotor. The maximum deviation between this angle and the actual angle is ±30°; after the motor rotates stably, 6 accurate Hall sector angles can be obtained within one electrical cycle; when calculating the rotational speed, use the signal capture function to trigger Interrupt, and read the counting results in the interrupt for speed calculation, clear the count at the same time, and enter the next sector count; the motor rotates for one electrical cycle, and Hall produces 6 sector changes; ω _e represents the electrical angular speed of the motor , the calculation formula is: In the formula, RegisterCnt is the read timer count value, f _timer is the frequency of the timer configuration; the mechanical speed is 1/P times the electrical angular speed, and the formula is: In the formula, P is the number of pole pairs of the motor, and ω _m represents the mechanical angular speed of the motor. 5. A control method for self-learning DC brushless motor Hall position based on vector algorithm according to claim 1, characterized in that: the final obtained θ _result is input to SVPWM, and Ud is equal to 0 through fixed Uq. It is equal to 10000 to generate a three-phase PWM wave that changes with time, and finally synthesizes a rotating vector to drive the motor rotor to rotate. 6. A control method for self-learning DC brushless motor Hall position based on vector algorithm according to claim 1, characterized in that: the host computer designs a UI interface, and operates the UI interface to communicate with the host computer and control panel. After receiving the corresponding instructions, the board performs the corresponding actions.