Preset super-twisting position sensor-less control method for ship propulsion motor

Abstract

Disclosed in the present invention is a preset super-twisting position sensor-less control method for a ship propulsion motor, belonging to the field of ship permanent magnet synchronous propulsion motor speed control. On the basis of traditional sliding mode observers, proposed in the present invention is an optimised super-twisting control algorithm with a time-varying gain and a switching function being introduced, realising fast response and low jitter of a control system. In addition, a preset time control is combined, ensuring that the system remains stable without being affected by other parameters or the initial state of the system. Moreover, a convergence time can be manually set according to requirements, and the convergence time of an angle error of a rotor of a ship propulsion motor can be controlled by means of setting a preset time T. Meanwhile, the rotor position estimated conditions at different preset times T are simulated. In the present invention, the preset super-twisting position sensor-less control method for a ship propulsion motor based on a preset super-twisting observer can improve the dynamic quality of a controlled system, and has significant advantages in terms of convergence speed and control accuracy.

Description

A sensorless control method for a ship propulsion motor with a pre-set super-spiral mechanism Technical Field

[0001] This invention relates to the field of ship propulsion motor control, and in particular to a sensorless control method for ship propulsion motors based on a preset super-helical observer. Background Technology

[0002] Permanent magnet synchronous motors (PMSMs), as a type of high-efficiency, low-energy-consumption motor, have significant applications in electrically propelled ships. Sensorless control technology estimates rotor position and speed by real-time monitoring of stator voltage and current during motor operation, eliminating traditional mechanical sensors. This not only saves space and cost in the motor control system but also improves its reliability. Traditional sliding mode control offers fast dynamic response and strong robustness. While high switching gain ensures system stability and improves anti-interference capabilities, it can lead to discontinuous control signals and severe chattering. Combining it with adaptive algorithms or phase-locked loops results in complex system structures, and the convergence time is affected by system state. Furthermore, compared to ordinary land-based motors, ship propulsion motors operate in harsher environments with more frequent changes in operating conditions, and are affected by numerous internal and external disturbances. Traditional sliding mode controllers and convergence tracking algorithms cannot meet practical requirements. Therefore, new control methods are needed to meet the higher requirements for estimation accuracy and convergence time in motor control systems.

[0003] Patent CN116208040A discloses a speed adaptive observation method based on full-order terminal sliding mode. This control method solves the problems of slow observation speed and large chattering in permanent magnet synchronous motors at high speeds. However, the design of the full-order terminal sliding mode observer is relatively complex, and the finite-time stability of the system is related to the system parameters. Patent CN114528669A discloses a design method for a super-helical controller for permanent magnet synchronous motors based on preset time stability. It achieves system stability without being affected by motor parameters, but does not apply the super-helical algorithm to the observer for rotor position estimation.

[0004] Summary of the Invention

[0005] The purpose of this invention is to provide a preset superhelical sensorless control method for ship propulsion motors. This method combines the principle of superhelical sliding mode observer with preset time control technology, which can realize the convergence time of the error between the estimated value and the actual value of the motor rotor by artificial control, so that the convergence time is not affected by other parameters in the system, thus realizing relatively accurate and fast sensorless control of permanent magnet synchronous motors.

[0006] This invention describes a sensorless control method for a ship propulsion motor with a pre-set super-helical mechanism, comprising:

[0007] Dynamic modeling was used to establish the system state equations of the permanent magnet synchronous motor.

[0008] A switching function is introduced to optimize the structure of a novel superspiral sliding mode observer and to improve the influence of the time-varying function ρ on the preset time T.

[0009] By combining the system state equation of the permanent magnet synchronous motor and the novel super-helical sliding mode observer, the errors of the observed current and observed back electromotive force of the novel super-helical sliding mode observer along the α-axis and β-axis are obtained respectively.

[0010] Design a current sliding surface with the current error as the independent variable;

[0011] The relevant information of the rotor is estimated, and the phase is compensated. The relevant information of the rotor includes the position information of the motor rotor and the estimated value of the rotational speed.

[0012] In some embodiments, the novel superspiral sliding mode observer consists of two second-order sliding mode observers.

[0013] In some implementations, the two second-order sliding mode observers are a stator current sliding mode observer and a back electromotive force sliding mode observer, respectively.

[0014] In some implementations, the system state equation of the permanent magnet synchronous motor is:

[0015] In the formula, u α u β i represents the stator voltage component in the stationary coordinate system. α i β R represents the stator current component in the stationary coordinate system. s L is the stator resistance. d For stator inductance, p is the differential operator, e α e β Ψ is the back potential component in the stationary coordinate system. f ω represents the flux linkage of a permanent magnet. e θ represents the rotor electrical angle of the motor. e This represents the electrical angle of the motor rotor. p is the differential operator. The estimated current in the αβ coordinate system; ρ is the estimated back electromotive force in the αβ coordinate system; k1 and k2 are gain coefficients; ρ is the time-varying coefficient.

[0016] In some implementations, after the dynamic modeling is completed, and stability is ensured over a preset time, the Lyapunov equations are constructed. These Lyapunov equations are as follows:

[0017] And prove

[0018] In some implementations, the equations for optimizing the novel superspiral sliding mode observer structure by introducing a switching function are as follows:

[0019] In the formula, k t tanh(k) is the switching gain constant. v x) is the switching function added to the current observer, a time-varying function.

[0020] In some implementations, the errors in the observed current and observed back electromotive force of the novel superspiral sliding mode observer along the α-axis and β-axis are:

[0021] In some implementations, the current sliding surface is:

[0022] In some implementations, the estimated rotor position information and estimated rotational speed are:

[0023] In some implementations, when the rotor position information has a phase lag, the estimated rotor position information needs to be compensated. This compensation is as follows:

[0024] In the formula, ω c This is the cutoff frequency of the low-pass filter.

[0025] The advantages of the pre-set super-spiral sensorless control method for ship propulsion motors described in this invention are as follows:

[0026] 1. The convergence time of the ship's propulsion motor rotor angle error can be controlled by setting a preset time T, and the convergence performance is not affected by other system parameters;

[0027] 2. It is proposed to introduce preset time performance into super-helical control, and combine it with sensorless control algorithm to effectively improve the motor rotor estimation accuracy and system reliability;

[0028] 3. Compared with finite-time stability, the improved superspiral controller based on preset time can ensure that the stability of the system is not affected by other parameters and the initial state of the system, and the convergence time can be set manually according to the system requirements, while also ensuring the low chattering characteristics of the system.

[0029] 4. The rotor position information estimated by the algorithm described in this invention has higher accuracy, faster convergence speed, and more stable speed and torque changes, which meets the requirements of ship three-phase permanent magnet synchronous propulsion motors for use in harsh underwater environments. Attached Figure Description

[0030] Figure 1 is a block diagram of a preset super-helical sensorless control system applied to a propulsion motor in some embodiments of the present invention;

[0031] Figure 2 shows the simulation results of the preset time algorithm controller in scenario 1 in some embodiments of the present invention;

[0032] Figure 3 shows the simulation results of the preset time algorithm controller in scenario 2 in some embodiments of the present invention;

[0033] Figure 4 shows the simulation results of the preset time algorithm controller in scenario 3 in some embodiments of the present invention;

[0034] Figure 5 shows the rotor observation position and error waveform of the T=0.001 preset super spiral algorithm in some embodiments of the present invention;

[0035] Figure 6 is a magnified view of the rotor error of the preset super spiral algorithm with T=0.001 in some embodiments of the present invention;

[0036] Figure 7 shows the rotor observation position and error waveform of the traditional sliding mode algorithm in some embodiments of the present invention;

[0037] Figure 8 is a magnified view of the rotor observation error of the traditional sliding mode algorithm in some embodiments of the present invention;

[0038] Figure 9 shows the rotor observation position and error waveform of the preset super spiral algorithm with T=0.002 in some embodiments of the present invention;

[0039] Figure 10 shows the rotor observation position and error waveform of the T=0.003 preset super spiral algorithm in some embodiments of the present invention. Detailed Implementation

[0040] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. The content and drawings described below are only some embodiments recorded in this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0041] To make the technical solutions and advantages of the present invention clearer, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention.

[0042] The state equation of the permanent magnet synchronous motor in the two-phase stationary coordinate system (α, β) is as follows:

[0043] In the formula: u α u βi represents the stator voltage component in the stationary coordinate system. α i β R represents the stator current component in the stationary coordinate system. s L is the stator resistance. d For stator inductance, p is the differential operator, e α e β Let ω be the back potential component in the stationary coordinate system. e θ represents the rotor electrical angle of the motor. e This indicates the electrical angle of the motor rotor.

[0044] Introducing the concept of a preset time, we construct a time-varying piecewise function:

[0045] In the formula, ε>0, λ>1, and T is the preset steady-state time constant. It is assumed that there exists a continuously differentiable positive definite function V(t) that satisfies the following conditions:

[0046] Find:

[0047] The system is preset to be stable over a certain time.

[0048] The state equation of the novel superspiral algorithm incorporating time-varying gain is as follows:

[0049] Where k1, k2, and k3 are gain coefficients, and δ1 and δ2 are disturbance terms.

[0050] To verify the convergence of the preset performance, the stability of the convergence of the state equation in equation (5) is first verified by ignoring the perturbation term, and the following Lyapunov function is selected:

[0051] Equation (6) simplifies to: V(x)=ξ T P0ξ(7)

[0052] In the formula,

[0053] When all k2 are greater than 0, P0 is a positive definite matrix, and V is a continuous positive definite function, thus we can obtain European-style space The 2-norm. Assume the maximum and minimum values ​​of the eigenvalues ​​of P0 are respectively... Then V satisfies

[0054] Differentiating both sides of equation (7), we get:

[0055] In the formula:

[0056] When both k1 and k2 are greater than 0, then P1 is a positive definite matrix. It is a negative definite value. Assume the minimum value of P1 is... Then there is

[0057] From equation (11), we can obtain

[0058] Combining (10) and (11), we can deduce

[0059] Therefore, according to Lyapunov's asymptotic stability theory, the system is asymptotically stable. The next step is to prove the algorithm's pre-defined time stability and calculate the values ​​of relevant parameters.

[0060] Construct matrix P2, and let P2 = P1 - P0, then we have

[0061] When both the first-order principal minor and the second-order principal minor of P2 are greater than zero, that is, when the following inequality is satisfied...

[0062] When appropriate values ​​of k1, k2, and k3 are chosen to satisfy inequality (15), P2 is a positive definite matrix, and thus: ρζ T P2ζ=ρζ T (P1-P0)ζ>0(16)

[0063] V(x) = ζ T P0ζ, Combining with equation (16), we get

[0064] The system has preset time stability, and the algorithm controller shown in equation (5) is built using the Matlab platform.

[0065] Setting up scenario 1: x1(0) = 10, x2(0) = 6, preset time T = 1, the response curves of x1 and x2 obtained after simulation are shown in Figure 2;

[0066] Setting up scenario 2: x1(0) = 10, x2(0) = 6, preset time T = 0.6, the response curves of x1 and x2 obtained after simulation are shown in Figure 3;

[0067] Scenario 3: x1(0) = 5, x2(0) = 8, preset time T = 1. The response curves of x1 and x2 obtained after simulation are shown in Figure 4. Other gain parameters k1 = 1, k2 = 1, k3 = 3, and the simulation step size is 1*10. -3 s.

[0068] By comparing Figure 2 and Figure 3, it can be seen that the state variables x1 and x2 can converge to 0 and remain stable before the preset time T, which is 1s and 0.5s respectively. This shows that in this new super-spiral sliding mode algorithm, the preset time T can be a parameter given by humans.

[0069] By comparing Figure 2 and Figure 4, under different initial values, the state variables x1 and x2 can converge to 0 and remain stable within the preset time T. This shows that the novel super-spiral sliding mode algorithm can converge and remain stable within the preset time T under different initial values.

[0070] To achieve the preset time performance and chattering suppression of the sliding mode observer, a novel superspiral sliding mode observer is proposed by combining the aforementioned novel preset superspiral algorithm.

[0071] The novel super-helical sliding mode observer consists of two second-order sliding mode observers. By combining equations (1) and (5), a stator current sliding mode observer and a back electromotive force sliding mode observer are constructed:

[0072] Where p is the differential operator, The estimated current is given in the αβ coordinate system. ρ(t) represents the estimated back electromotive force in the αβ coordinate system; k1 and k2 are gain coefficients; ρ(t) is a time-varying coefficient.

[0073] Once the preset time T is reached, the time-varying function ρ(t) becomes constant. This is because once the observer enters the sliding surface and stabilizes... This will become a smaller value, and the response time of the current observer will be relatively long, thus affecting the stability of the system; when encountering disturbances... The value of can suddenly increase, causing system chattering. Therefore, the observer structure can be optimized by introducing a switching function:

[0074] In the formula, k t The switching gain constant is determined by adding a switching function tanh(k) to the current observer. v x), on the one hand, can When the sliding mode is relatively small, appropriately increasing it can improve the observer's response speed; on the other hand, it can reduce the impact of disturbances. Excessive values ​​lead to additional chattering in the system. Compared to the traditional sign function sign(x), the switching function tanh(k)... v x) allows the system state to pass through the sliding surface at a low speed, thereby reducing observation errors caused by chattering. Furthermore, k can be adjusted... v The value of is used to control the saturation level of the hyperbolic tangent curve.

[0075] Taking the α-axis as an example, combining equation (1) and the observer (19), multiplying both sides of the equation in equation (1) by the time variable ρ in formula (2), and subtracting the result from the observer (19), we obtain the dynamic equation of the observation current and the error of the observation back electromotive force of the new super-spiral sliding mode observer.

[0076] Similarly, the dynamic equations for the observed β-axis current and the error in the observed back electromotive force can be obtained:

[0077] The improved observer consists of a current sliding mode observer and an extended back EMF sliding mode observer. The new current sliding surface is defined as:

[0078] Choose the current Lyapunov function as:

[0079] From the second method of Lyapunov functions and the reachability condition of sliding mode motion, it can be seen that when V i When V is positive time, i derivative It needs to be guaranteed to be a negative definite value, therefore it needs to satisfy the following condition.

[0080] Equation (23) must hold if the following conditions are met.

[0081] We can obtain:

[0082] In the formula, When the stator current reaches the sliding surface, the observed current value will converge to the actual value, that is... and p is a differential operator. When k1 > max{|e α |,|e β When |}, equation (26) can be satisfied, and it can be determined that the system is stable.

[0083] Based on the back EMF observer (19) of the novel super-helical sliding mode observer, the back EMF of the motor can be obtained through an integrator, and then the position and speed of the motor rotor can be further estimated. The estimated values ​​of the motor rotor position and speed can be obtained using equation (1):

[0084] The use of a low-pass filter will cause a phase lag in the motor rotor position information, so compensation is needed for the estimated rotor position.

[0085] In the formula, ω c This is the cutoff frequency of the low-pass filter.

[0086] Finally, the ship propulsion permanent magnet synchronous motor control system was modeled and simulated. The system model is shown in Figure 1. The motor parameter table is shown in Table 1, and the parameters of each module after debugging are shown in Table 2.

[0087] Table 1 PSPM Nominal Parameter Values

[0088] Table 2. Parameter values ​​of each module in the simulation.

[0089] To analyze the preset time performance of the novel super-helical sliding mode observer, the preset time T was set to 0.001, 0.002 and 0.003 respectively, and the observed rotor position information was compared with that of the traditional sliding mode observer. The simulation results are shown in Figure 5-10.

[0090] By comparing Figures 5 and 7, it can be seen that when a suitable preset time T is set, the preset superhelical algorithm observer has a faster convergence speed than the traditional sliding mode observer. The rotor angle error converges to 0 at time T = 0.0011s, while the traditional sliding mode observer only achieves convergence to 0 at simulation time T = 0.0014s after multiple parameter adjustments. To compare error fluctuations, Figure 6 is a partial enlarged view of Figure 5, and Figure 8 is a partial enlarged view of Figure 7. Figure 6 shows that the rotor error curve of the preset superhelical algorithm observer is relatively smooth with no obvious chattering, and the angle error is approximately 0.02 rad. Figure 8 shows that the chattering of the traditional sliding mode observer is more obvious, with an amplitude of approximately 0.005 rad and an average angle error of approximately 0.042 rad. Comparing Figures 6 and 8 demonstrates that the preset superhelical algorithm observer has less chattering and a smaller observation error compared to the traditional sliding mode observer, resulting in higher observation accuracy.

[0091] Comparing Figures 5, 9, and 10 shows that under the control of the new super-helical sliding mode observer, even with different preset times T, the system can achieve high-precision estimation of the rotor position within the preset time T.

[0092] All of the above-mentioned undisclosed matters can be implemented using existing technologies, so they will not be elaborated here.

[0093] For those skilled in the art, several similar modifications and improvements can be made without departing from the inventive concept of this invention, and these should also be considered within the scope of protection of this invention.

Claims

1. A sensorless control method for a ship propulsion motor with a pre-set superspiral mechanism, characterized in that, The pre-set super-helical sensorless control method for ship propulsion motors includes: Dynamic modeling was used to establish the system state equations of the permanent magnet synchronous motor. A switching function is introduced to optimize the structure of a novel superspiral sliding mode observer and to improve the influence of the time-varying function ρ on the preset time T. By combining the system state equation of the permanent magnet synchronous motor and the novel super-helical sliding mode observer, the errors of the observed current and observed back electromotive force of the novel super-helical sliding mode observer along the α-axis and β-axis are obtained respectively. Design a current sliding surface with the current error as the independent variable; The relevant information of the rotor is estimated, and the phase is compensated. The relevant information of the rotor includes the position information of the motor rotor and the estimated value of the rotational speed.

2. The method for sensorless control of a ship propulsion motor with a pre-set superhelical configuration according to claim 1, wherein the novel superhelical sliding mode observer is composed of two second-order sliding mode observers.

3. The method for sensorless control of a ship propulsion motor with a pre-set super-spiral mechanism according to claim 2, wherein, The two second-order sliding mode observers are the stator current sliding mode observer and the back electromotive force sliding mode observer.

4. The method for sensorless control of a ship propulsion motor with a pre-set super-spiral mechanism according to claim 3, wherein, The system state equation of the permanent magnet synchronous motor is: In the formula, u α u β i represents the stator voltage component in the stationary coordinate system. α i β R represents the stator current component in the stationary coordinate system. s L is the stator resistance. d For stator inductance, p is the differential operator, e α e β Ψ is the back potential component in the stationary coordinate system. f ω represents the flux linkage of a permanent magnet. ε θ represents the rotor electrical angle of the motor. ε This represents the electrical angle of the motor rotor. p is the differential operator. The estimated current in the αβ coordinate system; ρ is the estimated back electromotive force in the αβ coordinate system; k1 and k2 are gain coefficients; ρ is the time-varying coefficient.

5. A sensorless control method for a pre-set super-spiral propulsion motor of a ship according to claim 4, wherein, After the dynamic modeling is completed, and stability is ensured at a preset time, the Lyapunov equations are constructed as follows: And prove 6. A sensorless control method for a pre-set superspiral propulsion motor of a ship according to claim 4 or 5, wherein, The equation for optimizing the novel superspiral sliding mode observer structure by introducing a switching function is as follows: In the formula, k t tanh(k) is the switching gain constant. v x) is the switching function added to the current observer, a time-varying function.

7. The method for sensorless control of a ship propulsion motor with a pre-set superhelical axis according to claim 6, wherein the errors of the observed current and observed back electromotive force of the novel superhelical sliding mode observer along the α-axis and β-axis are:

8. A sensorless control method for a pre-set super-spiral propulsion motor of a ship according to claim 7, wherein, The current sliding surface is:

9. A sensorless control method for a pre-set super-spiral propulsion motor of a ship according to claim 8, wherein, The estimated rotor position and rotational speed are:

10. A sensorless control method for a pre-set super-spiral propulsion motor of a ship according to claim 9, wherein, When the rotor position information has a phase lag, the estimated rotor position information needs to be compensated. This compensation is as follows: In the formula, ω c This is the cutoff frequency of the low-pass filter.

Images