Current ripple suppression method of underwater propulsion motor

Abstract

The invention discloses a current ripple suppression method for an underwater propulsion motor, relates to the technical field of motor control, and reduces overshoot and adjustment time when the rotating speed of the motor changes suddenly by improving a traditional field-oriented control framework and adopting a differential tracker to smooth an input reference rotating speed. Dead-beat model-free predictive control and an ESO observer are introduced into a current loop, and current ripples caused by parameter mismatch and external interference are effectively restrained. Model-free predictive control is adopted, the number of mathematical model parameters is reduced, the model accuracy is improved, and the control precision is improved. The current ripple is further reduced by identifying the inductance parameter and adjusting the PWM frequency according to the actual inductance value. According to the invention, the overshoot phenomenon in the rotating speed adjusting process is obviously reduced, the stability and efficiency of motor operation are effectively improved, and the service life of the motor is prolonged. The invention provides a new solution for improving the performance of the underwater robot propulsion system.

Description

Technical Field

[0001] This invention relates to the field of motor control technology, and more specifically to a method for suppressing current ripple in an underwater propulsion motor. Background Technology

[0002] As an important tool for marine resource exploration and development, underwater robots not only need to adapt to various complex marine environments, but also need long-lasting and efficient endurance, which places higher demands on their electric propulsion systems.

[0003] Torque pulsation during motor operation is the main cause of vibration and noise in underwater propulsion. The generation of torque pulsation is directly related to the current ripple of the motor. Suppressing the current ripple can effectively reduce torque pulsation and noise during motor operation, while also reducing motor losses and temperature rise, improving efficiency and reliability, enhancing propulsion smoothness and control precision, thereby improving the overall reliability and lifespan of the system.

[0004] Current research on suppressing motor current ripple mainly focuses on three aspects: inverter current ripple research, dead time optimization, and control algorithm optimization. The following section will describe the current research status from the perspective of control algorithm optimization.

[0005] Active disturbance rejection control (ADRC) has attracted widespread attention in the field of permanent magnet synchronous motor (PMSM) control. Its intuitive mathematical description and easily analyzable transfer function properties are the reasons for its widespread use. However, traditional ADRC techniques are insufficient to fully address various interferences in motor systems. To enhance the harmonic suppression capability of ADRC, improvements have been made. Existing technologies propose adaptive ADRC to suppress uncertain current ripple. The extended state observer in ADRC is optimized by an adaptive resonant controller, thereby suppressing uncertain periodic and aperiodic interferences. The effectiveness of this method has been verified experimentally. Existing technologies also disclose a model-free, deadbeat predictive current control method based on an extended state observer, introducing adaptive gain. This method suppresses current ripple caused by controller gain adaptation, improving motor performance. Furthermore, by combining deadbeat current control (DPCC) with high-frequency current injection, existing technologies can observe the inductance, flux linkage, and resistance values ​​of the PMSM online. This method fully utilizes the high bandwidth and high parameter sensitivity of DPCC. Experiments show that even under severe parameter adaptation, the PMSM can still operate well.

[0006] However, many problems still exist in the aforementioned existing technologies:

[0007] When the motor speed changes abruptly, there is a large overshoot and a long adjustment time.

[0008] The mathematical model of an electric motor contains multiple parameters, such as inductance, resistance, and flux linkage, which can easily lead to inaccuracy in the mathematical model.

[0009] When the actual inductance parameter inside the motor decreases, adjusting the controller parameters cannot restore the three-phase current waveform of the motor to its original state, and the current ripple will increase.

[0010] Therefore, there is an urgent need for a method that can effectively suppress current ripple in underwater propulsion motors. Summary of the Invention

[0011] In view of this, the purpose of the present invention is to provide a method for suppressing current ripple in an underwater propulsion motor.

[0012] To achieve the above objectives, the present invention adopts the following technical solution:

[0013] This invention first provides a method for suppressing current ripple in an underwater propulsion motor, wherein the underwater propulsion motor is a permanent magnet synchronous motor based on field-oriented control, and includes the following steps:

[0014] S1: Receive the reference rotational speed from the external input, smooth the sudden speed changes in the reference rotational speed through the differential tracker, obtain a smooth reference rotational speed, and input it to the rotational speed loop;

[0015] S2: Real-time sampling of the actual rotational speed fed back by the permanent magnet synchronous motor at the current moment, calculation of the speed error between the smooth reference speed and the actual speed, and generation of d / q axis current reference values ​​based on the speed error through the PI controller;

[0016] S3: Real-time sampling of the actual d / q axis current fed back by the permanent magnet synchronous motor at the current moment, and obtaining the motor estimate through the ESO observer;

[0017] S4: The estimated value of the motor is used to compensate for the delay of the actual d / q axis current in the deadbeat-free model-free predictive control law for two sampling cycles. The deadbeat-free model-free predictive control law uses the reference value of the d / q axis current after two sampling cycles as the expectation of the predicted d / q axis current after two sampling cycles, and predicts and calculates the predicted d / q axis reference voltage value after one sampling cycle.

[0018] S5: Identify the actual inductance of the permanent magnet synchronous motor online, and generate PWM frequency adjustment commands based on the difference between the actual inductance and the nominal inductance;

[0019] S6: PWM drives the three-phase voltage of the permanent magnet synchronous motor and outputs torque according to the PWM frequency adjustment command and the d / q axis stator voltage control value.

[0020] Preferably, in step S1, a second-order differential tracker is used to process the input reference rotational speed, and the mathematical model is as follows:

[0021]

[0022] In the formula, v(k) is the reference rotational speed command at time k; z1 and z2 are two output signals, where z1 is the position tracking output for k sampling periods. z2 is the velocity tracking output for k sampling periods; T S is the sampling period of the current loop; r is the speed factor, which determines the tracking speed of the signal; h is the filtering factor, which filters the noise of the signal; fhan() is the fastest control synthesis function.

[0023] Preferably, the step in S3 of performing delay compensation for the deadbeat-free model-predictive control law based on the estimated value of the ESO observer for two sampling periods includes:

[0024] The actual d / q axis currents fed back by the permanent magnet synchronous motor at the current moment are sampled in real time to construct an ESO observer, and the estimated values ​​of d / q axis currents and unknown disturbances are updated in real time after one sampling period based on the actual d / q axis currents.

[0025] Preferably, the ESO observer is used to implement the following steps:

[0026]

[0027]

[0028] In the formula, i d i q These are the actual stator currents along the d and q axes, respectively. i d i q The estimated value; For unknown disturbance estimates; ε rrd ε rrq The error between the estimated value and the actual stator current; α0 is the reciprocal of the initial inductance of the permanent magnet synchronous motor; u d u q These are the d / q axis stator voltage control values, respectively.

[0029]

[0030] In the formula, ω b This represents the observer bandwidth.

[0031] Preferably, step S4 includes the following steps:

[0032] Based on the hyperlocal model, a deadbeat-free model-free predictive control law is constructed. Based on the delay compensation algorithm, the reference value of the d / q axis current after two sampling periods is assigned to the actual d / q axis current after two sampling periods. Combining the estimated value of the d / q axis current after one sampling period and the estimated value of the unknown disturbance, the predicted d / q axis reference voltage value after one sampling period is calculated.

[0033] Preferably, the deadbeat-free model-free predictive control law includes:

[0034]

[0035] Other The predicted d / q axis reference voltage value is expressed as:

[0036]

[0037] In the formula, α0 is the reciprocal of the initial inductance of the permanent magnet synchronous motor; T s For the current loop sampling period; F d (k), F q (k) represent the estimated perturbation values ​​of the d / q axes at time k; i d (k+2), i q (k+2) represent the predicted d / q axis currents after two sampling periods; i d (k+1),i q (k+1) represent the predicted d / q axis currents after one sampling period; u d (k+1), u q (k+1) represent the d / q axis stator voltage control values ​​after one sampling period; These are the d / q axis current reference values ​​after two sampling periods; These are the estimated values ​​of the unknown disturbance after one sampling period.

[0038] Preferably, the inductance parameter identification algorithm for identifying the actual inductance of the permanent magnet synchronous motor in S5 is a recursive least squares algorithm.

[0039] Preferably, in step S5, the difference between the actual inductance and the nominal inductance is compared to obtain a corresponding numerical relationship, and a PWM frequency adjustment command is generated based on the numerical relationship to change the PWM frequency; the numerical relationship is an integer relationship.

[0040] As can be seen from the above technical solution, compared with the prior art, the present invention discloses a method for suppressing current ripple of an underwater propulsion motor, which has the following beneficial effects:

[0041] This invention greatly suppresses the overshoot problem caused by sudden changes in motor speed. By adding a differential tracker after the input value, compared with using PI control alone, it can effectively reduce the overshoot caused by sudden changes in speed, enabling the underwater propulsion motor to smoothly adjust the speed, and the overshoot has a significant suppression effect.

[0042] This invention adds delay compensation to the extended state observer and uses an inductance parameter identification method to adjust the PWM frequency according to the ratio between the actual inductance and the controller inductance. Compared with the traditional method of observing and compensating for external interference, the strategy proposed in this invention is more effective and has a more obvious effect on suppressing current ripple.

[0043] This invention employs model-free predictive control, which reduces the number of mathematical model parameters, improves model accuracy, and enhances control precision.

[0044] By combining the above methods, the current ripple caused by the decrease in inductance parameters is effectively suppressed, ensuring the stable operation of the underwater thruster and improving the service life of the motor. Attached Figure Description

[0045] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on the provided drawings without creative effort.

[0046] Figure 1 A schematic diagram of the current ripple suppression method for underwater propulsion motors provided by the present invention;

[0047] Figure 2 This is a schematic diagram of the ESO computing structure provided by the present invention.

[0048] Figure 3 The schematic diagram of the deadbeat-free model-free current prediction control structure provided by the present invention;

[0049] Figure 4 The schematic diagram of the PWM control structure provided by this invention;

[0050] Figure 5 The MATLAB simulation interface diagram provided by this invention;

[0051] Figure 6 The motor speed diagram provided for this invention;

[0052] Figure 7 The present invention provides a conventional PI control motor speed diagram;

[0053] Figure 8The three-phase current waveform diagram of the motor provided for this invention;

[0054] Figure 9 This invention provides a conventional PI control for three-phase current.

[0055] Figure 10 This invention provides a three-phase current diagram of the unadjusted PWM frequency.

[0056] Figure 11 This is a spectrum analysis diagram of the unadjusted PWM frequency waveform provided by the present invention;

[0057] Figure 12 The three-phase current diagram for adjusting the PWM frequency provided by this invention;

[0058] Figure 13 The waveform spectrum analysis diagram for adjusting the PWM frequency provided by this invention. Detailed Implementation

[0059] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0060] This invention discloses a current ripple suppression method for an underwater propulsion motor. The underwater propulsion motor is a permanent magnet synchronous motor based on field-oriented control. The main control methods for this type of motor include direct torque control (DTC) and field-oriented control (FOC). The control method proposed in this invention is mainly an improvement within the control framework of FOC.

[0061] like Figure 1 The diagram shows the principle of the current ripple suppression method for underwater propulsion motors. It is based on vector control (FOC) and the main function of adding a tracking differentiator module after the input is to process the sudden changes in the input to make the change smoother and prevent the motor from generating excessive overshoot.

[0062] The speed loop uses PI control mainly for the purpose of simplifying control and reducing the computational load on the microcontroller.

[0063] The current loop uses deadbeat current predictive control based on a hyperlocal model. When used in conjunction with an extended observer (ESO), it can reduce parameter mismatch caused by too many parameters. Furthermore, the ESO can observe disturbances and compensate for the current loop output, thus effectively suppressing current ripple.

[0064] Specifically, the following steps are included:

[0065] S1: Receives the externally input reference speed and smooths abrupt changes in the reference speed using a differential tracker to obtain a smooth reference speed, which is then input to the speed loop. Specifically, a differential tracker (TD) is used to process the input reference speed. The differential tracker in the active disturbance rejection controller softens abrupt changes in the input speed, thereby reducing overshoot. This is because the motor is a nonlinear and time-varying system.

[0066] In one embodiment, a second-order differential tracker is used in S1 to process the input reference rotational speed. The mathematical model is as follows:

[0067]

[0068] In the formula, v(k) is the reference rotational speed command at time k; z1 and z2 are two output signals, where z1 is the position tracking output for k sampling periods. z2 is the velocity tracking output for k sampling periods; T S is the sampling period of the current loop; r is the speed factor, which determines the tracking speed of the signal; h is the filtering factor, which filters the noise of the signal; fhan() is the fastest control synthesis function.

[0069] The formula for the fastest tracking function, u = fhan(v(k), z1(k), z2(k), r, h), is as follows:

[0070]

[0071] After the input speed passes through the TD module, a smooth tracking signal is generated and input to the speed loop in the FOC control framework.

[0072] S2: Real-time sampling of the actual rotational speed fed back by the permanent magnet synchronous motor at the current moment, calculation of the speed error between the smoothed reference speed and the actual speed, and generation of d / q axis current reference values ​​based on the speed error using a PI controller. It should be noted that, to reduce the overall control complexity, PI control is used in the speed loop.

[0073] S3: Real-time sampling of the actual d / q axis current fed back by the permanent magnet synchronous motor at the current moment, and obtaining the motor estimate through the ESO observer.

[0074] It should be noted that deadbeat-free model-predictive control is used in the motor's current loop. To estimate the error, an extended state observer is used to observe the disturbances generated during motor operation. Furthermore, since deadbeat-free current predictive control is applied in the actual control process, at kT... S The motor current is sampled at constant times, at kT S to (k+1)T SThe PWM duty cycle signal is calculated within the sampling interval, therefore, in (k+1)T S The duty cycle signal is updated constantly, and it takes one cycle for the duty cycle signal to be applied to the inverter. Therefore, the total system delay is 2 sampling cycles. So, an algorithm with delay compensation is introduced here to offset the periodic delay that exists in the system at this time.

[0075] In one embodiment, the step in S3 of performing delay compensation for the deadbeat model-free predictive control law based on the ESO observer's estimate for two sampling periods includes:

[0076] The actual d / q axis currents fed back by the permanent magnet synchronous motor at the current moment are sampled in real time to construct an ESO observer, and the estimated values ​​of d / q axis currents and unknown disturbances are updated in real time after one sampling period based on the actual d / q axis currents.

[0077] In this embodiment, i is selected. d i q ;F d F q The extended state observer (ESO) is constructed using state variables. The ESO observer is used to implement the following steps:

[0078]

[0079] In the formula, i d i q These are the actual stator currents along the d and q axes, respectively. i d i q The estimated value; For unknown disturbance estimates; ε rrd ε rrq The error between the estimated value and the actual stator current; α0 is the reciprocal of the initial inductance of the permanent magnet synchronous motor; u d u q These are the d / q axis stator voltage control values, respectively.

[0080]

[0081] In the formula, ω b The observer bandwidth determines the dynamic performance of the current loop.

[0082] like Figure 2 The diagram shows the ESO computational structure. In this embodiment, after the ESO observer is discretized, the i of the next sampling period can be obtained. d i q The estimated value and the estimated value of the unknown disturbance in the next sampling period. The predicted current value at the next moment is used as the input to the current predictive control, and the estimated disturbance value is used as the compensation value. The introduction of an ESO observer can estimate unknown disturbances and reduce current ripple by compensating for them.

[0083] S4: The actual d / q axis current in the deadbeat-free model-free predictive control law is delayed by two sampling cycles using the motor estimate. The deadbeat-free model-free predictive control law uses the d / q axis current reference value after two sampling cycles as the expected value of the predicted d / q axis current after two sampling cycles, and predicts and calculates the predicted d / q axis reference voltage value after one sampling cycle.

[0084] In one embodiment, S4 includes the following steps:

[0085] Based on the hyperlocal model, a deadbeat-free model-free predictive control law is constructed. Based on the delay compensation algorithm, the reference value of the d / q axis current after two sampling periods is assigned to the actual d / q axis current after two sampling periods. Combining the estimated value of the d / q axis current after one sampling period and the estimated value of the unknown disturbance, the predicted d / q axis reference voltage value after one sampling period is calculated.

[0086] In this embodiment, the mathematical model under the traditional d / q two-phase rotating coordinate system is as follows:

[0087]

[0088] In the formula:

[0089] u d u q These are the stator voltages along the d-axis and q-axis;

[0090] i d i q These are the stator currents along the d-axis and q-axis;

[0091] For the stator flux linkages along the d-axis and q-axis;

[0092] R S Stator resistance;

[0093] ω e It represents the electric angular velocity.

[0094] Based on this, and according to the hyperlocal model, the mathematical model for deadbeat-free model-free predictive control is as follows:

[0095]

[0096] According to the above formula, the discrete form of the prediction model can be expressed as:

[0097]

[0098] Therefore, a delay compensation algorithm is introduced. The deadbeat-free model-predictive control law in this embodiment includes:

[0099]

[0100] Other The predicted d / q axis reference voltage value is expressed as:

[0101]

[0102] In the formula, α0 is the reciprocal of the initial inductance of the permanent magnet synchronous motor; T s For the current loop sampling period; F d (k), F q (k) represent the estimated perturbation values ​​of the d / q axes at time k; i d (k+2), i q (k+2) represent the predicted d / q axis currents after two sampling periods; i d (k+1),i q (k+1) represent the predicted d / q axis currents after one sampling period; u d (k+1), u q (k+1) represent the d / q axis stator voltage control values ​​after one sampling period; These are the d / q axis current reference values ​​after two sampling periods; These are the estimated values ​​of the unknown disturbance after one sampling period.

[0103] The disturbance estimates may include: disturbances caused by parameter mismatch (Ls, Rs drift), unmodeled dynamic disturbances, and external disturbances.

[0104] like Figure 3 The diagram shows the deadbeat-free model-free current predictive control architecture. This diagram represents the overall control scheme for the current loop, which is achieved by combining strategies such as delay compensation and extended state observers. The current loop employs deadbeat current predictive control based on a hyperlocal model. By combining it with an extended state observer (ESO), parameter mismatch caused by excessive parameters can be reduced. Furthermore, the ESO can observe disturbances and compensate for the current loop output, effectively suppressing current ripple.

[0105] S5: Identifies the actual inductance of the permanent magnet synchronous motor online and generates a PWM frequency adjustment command based on the difference between the actual and nominal inductance. It should be noted that prolonged motor operation can lead to a decrease in the motor's internal inductance parameters, resulting in increased current ripple at the motor's output. To suppress this current ripple caused by the decrease in actual parameters, the PWM frequency is adjusted.

[0106] In one embodiment, the inductance parameter identification algorithm for identifying the actual inductance of the permanent magnet synchronous motor online in S5 is a recursive least squares algorithm.

[0107] In one embodiment, in step S5, the difference between the actual inductance and the nominal inductance is compared to obtain a corresponding numerical relationship. A PWM frequency adjustment command is then generated based on this numerical relationship to change the PWM frequency. For ease of microcontroller processing, the numerical relationship is an integer relationship. By comparing the controller inductance value with the identified actual inductance value, the PWM frequency is adjusted, thereby increasing the switching frequency of the motor's internal power transistor (IGBT). This method can significantly suppress current ripple caused by a decrease in the actual inductance parameter.

[0108] S6: PWM drives the three-phase voltage of the permanent magnet synchronous motor and outputs torque according to the PWM frequency adjustment command and the d / q axis stator voltage control value.

[0109] like Figure 4 As shown, the FOC control framework specifically uses the SVPWM module: u d u q u is obtained after inverse Park transformation α u β By decoupling u α u β PWM is generated to control the IGBTs in the inverter, thereby generating torque for the motor.

[0110] like Figure 5 The image shown is a screenshot of the MATLAB simulation interface. The following MATLAB simulation will compare the differences between this invention and traditional PI control:

[0111] The motor parameters are set as follows: number of pole pairs Pn = 4; inductance Ls = 0.0185H; resistance R = 2.875Ω; flux linkage φ = 0.175. For example... Figure 6 The figure shown is a diagram of motor speed in an embodiment of the present invention. Figure 7 The traditional PI-controlled motor speed diagram provided by this invention shows that the horizontal axis represents time (seconds) and the vertical axis represents speed (r / min). (Comparison) Figure 6 and Figure 7 It can be seen that traditional PI control will produce overshoot when the speed changes abruptly, while the method of the present invention has almost no overshoot when the speed changes abruptly, and the speed transition is more stable.

[0112] like Figure 8 The image shows the three-phase current waveforms of the motor. Figure 9 This is a traditional PI control three-phase current diagram. The horizontal axis represents time (seconds), and the vertical axis represents current (A). Comparison Figure 8 and Figure 9It can be concluded that the traditional PI-controlled three-phase current has a large current ripple in steady state, while the control method proposed in this invention has a more stable waveform and better performance.

[0113] When the inductance parameter Ls = 0.0185 drops to 0.0085, the parameter mismatch is analyzed by comparing the differences between adjusting the PWM frequency and not adjusting the PWM frequency through simulation.

[0114] like Figure 10 The diagram shown is the three-phase current diagram without PWM frequency adjustment. Figure 11 The image shows the spectrum analysis of the unadjusted PWM frequency waveform. THD = 16.54%, which reflects the ripple level of the current waveform when the motor is in steady state. The smaller the value, the less ripple the waveform and the more stable the current.

[0115] like Figure 12 The diagram shown is a three-phase current diagram for adjusting the PWM frequency. Figure 13 The image shows a spectrum analysis of the PWM frequency waveform. THD = 9.62%. (This is achieved through comparison.) Figure 11 and Figure 13 The THD value of the response indicates that adjusting the PWM using the parameter identification method can effectively reduce the THD value, thus effectively suppressing current ripple.

[0116] Through the above comparison, the control method based on the present invention can effectively control the problems of motor speed overshoot caused by sudden speed changes, parameter mismatch, and increased current ripple caused by external interference.

[0117] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on its differences from other embodiments. Similar or identical parts between embodiments can be referred to interchangeably. For the apparatus disclosed in the embodiments, since they correspond to the methods disclosed in the embodiments, the description is relatively simple; relevant parts can be referred to the method section.

[0118] The above description of the disclosed embodiments enables those skilled in the art to make or use the invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the invention. Therefore, the invention is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for suppressing current ripple in an underwater propulsion motor, wherein the underwater propulsion motor is a permanent magnet synchronous motor based on field-oriented control, characterized in that, Includes the following steps: S1: Receive the reference rotational speed from the external input, smooth the sudden speed changes in the reference rotational speed through the differential tracker, obtain a smooth reference rotational speed, and input it to the rotational speed loop; S2: Real-time sampling of the actual rotational speed fed back by the permanent magnet synchronous motor at the current moment, calculation of the speed error between the smooth reference speed and the actual speed, and generation of d / q axis current reference values ​​based on the speed error through the PI controller; S3: Real-time sampling of the actual d / q axis current fed back by the permanent magnet synchronous motor at the current moment, and obtaining the motor estimate through the ESO observer; S4: The estimated value of the motor is used to compensate for the delay of the actual d / q axis current in the deadbeat-free model-free predictive control law for two sampling cycles. The deadbeat-free model-free predictive control law uses the reference value of the d / q axis current after two sampling cycles as the expectation of the predicted d / q axis current after two sampling cycles, and predicts and calculates the predicted d / q axis reference voltage value after one sampling cycle. S5: Identify the actual inductance of the permanent magnet synchronous motor online, and generate PWM frequency adjustment commands based on the difference between the actual inductance and the nominal inductance; S6: PWM drives the three-phase voltage of the permanent magnet synchronous motor and outputs torque according to the PWM frequency adjustment command and the d / q axis stator voltage control value.

2. The current ripple suppression method for an underwater propulsion motor according to claim 1, characterized in that, In S1, a second-order differential tracker is used to process the input reference rotation speed. The mathematical model is as follows: In the formula, v(k) is the reference rotational speed command at time k; z1 and z2 are two output signals, where z1 is the position tracking output for k sampling periods. z2 is the velocity tracking output at time k; T S is the sampling period of the current loop; r is the speed factor, which determines the tracking speed of the signal; h is the filtering factor, which filters the noise of the signal; fhan() is the fastest control synthesis function.

3. The current ripple suppression method for an underwater propulsion motor according to claim 1, characterized in that, The step in S3, which compensates for the delay of the deadbeat model-free predictive control law based on the ESO observer's estimate for two sampling periods, includes: The actual d / q axis currents fed back by the permanent magnet synchronous motor at the current moment are sampled in real time to construct an ESO observer, and the estimated values ​​of d / q axis currents and unknown disturbances are updated in real time after one sampling period based on the actual d / q axis currents.

4. The current ripple suppression method for an underwater propulsion motor according to claim 3, characterized in that, The ESO observer is used to perform the following steps: In the formula, i d i q These are the actual stator currents along the d and q axes, respectively. i d i q The estimated value; For unknown disturbance estimates; ε rrd ε rrq The error between the estimated value and the actual stator current; α0 is the reciprocal of the initial inductance of the permanent magnet synchronous motor; u d u q These are the d / q axis stator voltage control values, respectively. In the formula, ω b This represents the observer bandwidth.

5. The current ripple suppression method for an underwater propulsion motor according to claim 1, characterized in that, S4 includes the following steps: Based on the hyperlocal model, a deadbeat-free model-free predictive control law is constructed. Based on the delay compensation algorithm, the reference value of the d / q axis current after two sampling periods is assigned to the actual d / q axis current after two sampling periods. Combining the estimated value of the d / q axis current after one sampling period and the estimated value of the unknown disturbance, the predicted d / q axis reference voltage value after one sampling period is calculated.

6. The current ripple suppression method for an underwater propulsion motor according to claim 5, characterized in that, The deadbeat-free model-free predictive control law includes: Other The predicted d / q axis reference voltage value is expressed as: In the formula, α0 is the reciprocal of the initial inductance of the permanent magnet synchronous motor; T s For the current loop sampling period; F d (k), F q (k) represent the estimated perturbation values ​​of the d / q axes at time k; i d (k+2), i q (k+2) represent the predicted d / q axis currents after two sampling periods; i d (k+1),i q (k+1) represent the predicted d / q axis currents after one sampling period; u d (k)+1), u q (k+1) represent the d / q axis stator voltage control values ​​after one sampling period; These are the d / q axis current reference values ​​after two sampling periods; These are the estimated values ​​of the unknown disturbance after one sampling period.

7. The current ripple suppression method for an underwater propulsion motor according to claim 1, characterized in that, The inductance parameter identification algorithm for online identification of the actual inductance of the permanent magnet synchronous motor in S5 is a recursive least squares algorithm.

8. The current ripple suppression method for an underwater propulsion motor according to claim 1, characterized in that, In step S5, the difference between the actual inductance and the nominal inductance is compared to obtain the corresponding numerical relationship. The PWM frequency adjustment command is generated through the numerical relationship to change the PWM frequency. The numerical relationship is an integer relationship.

Images