阅读说明：本技术 一种混合励磁电机弱磁区稳态效率寻优控制方法 (Mixed excitation motor weak magnetic region steady state efficiency optimization control method ) 是由 樊英 陆星池 雷宇通 于 2021-08-18 设计创作，主要内容包括：本发明公开了一种混合励磁电机弱磁区稳态效率寻优控制方法,该方法基于新型交替极混合励磁电机数学模型,推导出了电机的损耗模型,并基于该损耗模型和寻优过程中电机电磁特性的分析,提出一种基于输入功率梯度法和弱磁裕度相结合的d轴磁链扰动观测寻优策略。与现有技术相比,本发明在观测输入功率的同时,对其梯度进行标幺化处理,并作为步长系数,在提高系统效率的同时实现变步长寻优。除此之外,本发明还通过对弱磁裕度的定义,进一步解决了现有变步长效率寻优技术中存在的初始步长确定困难这一问题,并与寻优算法结合避免了某些工况下弱磁失控的可能性。(The invention discloses a hybrid excitation motor weak magnetic area steady state efficiency optimizing control method, which is based on a novel alternate pole hybrid excitation motor mathematical model, deduces a loss model of a motor, and provides a d-axis flux linkage disturbance observation optimizing strategy based on the combination of an input power gradient method and weak magnetic margin based on the loss model and the analysis of the electromagnetic characteristics of the motor in the optimizing process. Compared with the prior art, the method provided by the invention can observe the input power, perform per unit processing on the gradient of the input power, and realize variable step length optimization while improving the system efficiency as a step length coefficient. In addition, the invention further solves the problem of difficult initial step size determination in the prior variable step size efficiency optimizing technology by defining the weak magnetic margin, and avoids the possibility of the out-of-control weak magnetic under certain working conditions by combining with an optimizing algorithm.)

1. A hybrid excitation motor weak magnetic area steady state efficiency optimizing control method is characterized by comprising the following steps:

firstly, before optimization is started, phase current, torque and rotating speed signals of an inner stator and an outer stator of a motor at the current moment (k moment) are acquired in real time through sampling, and current (i) of the motor under a synchronous rotating coordinate system is obtained through coordinate transformation (Clarke transformation and Park transformation) and a PI regulator_dq(k)、i_dqi(k) And a command voltage (v)_dq(k)、v_dqi(k))；

Secondly, the sampled current (i)_dq(k)、i_dqi(k) And a command voltage (v)_dq(k)、v_dqi(k) Psi is obtained via a power, flux linkage and weak magnetic margin calculation module, in addition to generating a switching signal by an SVPWM module to power the motor_d(k)、P_e(k) And delta_vA real-time value of (c);

the derivation of the d-axis flux linkage characteristics based on the loss model, after the optimization start signal is given, except that the current time psi is calculated in real time_d(k)、P_e(k) And delta_vIn addition to the above value, the current optimization method also compares the current optimization method with the corresponding parameter of the previous period, obtains the change direction and the change size of the d-axis flux linkage step length of the next optimization period through a variable step length optimization logic module, and finally changes the current set value through a current distribution module in real time to realize the transition from the working point of the motor to the efficiency optimum point, wherein the specific optimization process comprises the following steps:

(1) in the first period of optimization, the change direction of the magnetic linkage of the d axis, which improves the efficiency of the motor, is found out tentatively in a given step length mode, and the initial calculation power is stored for doingIs a base value P_basisMeanwhile, the variation delta of the weak magnetic margin in the fixed-step tentative optimization process_vOf combined typeObtaining the value of alpha, and establishing a relation between the weak magnetic margin and the d-axis flux linkage;

(2) in the second period of the optimization, the calculated value delta based on the weak magnetic margin_vAnd α, according to the formula Δ ψ_{d_init}＝0.618×δ_vDetermining a variable step length optimization initial value so as to improve the convergence speed and ensure the stability of the motor in the optimization process;

(3) starting from the third period of the seek, at Δ ψ_{d_init}Determining the flux linkage step length of each optimization period as initial step length by using the per-unit input power gradient value as step length coefficient to realize rapid convergence, i.e.Δψ_{d_vs}＝Grad_Pe*×Δψ_{d_init}；

(4) When the efficiency optimal point is approached, the algorithm is automatically switched to a Rosenbrock strategy, so that the convergence rate is ensured, and meanwhile, the convergence is ensured.

2. The method as claimed in claim 1, wherein the derivation of the loss characteristic of the novel hybrid excitation motor based on the d-axis flux linkage, the efficiency optimization strategy based on disturbance observation, the flux linkage step selection mode combined with per unit input power gradient, the initial step selection formula based on weak magnetic margin, the current distribution mode based on the motor torque flux linkage model, and the modified Rosenbrock algorithm combined with weak magnetic margin are performed.

3. The method of claim 2, wherein the loss characteristic of the novel hybrid excitation motor based on the d-axis flux linkage is derived as follows:

novel alternating pole hybrid excitation motor relates to d and q axis magnetic linkage psi_d、ψ_qControllable ofThe loss (copper, iron, stray losses) is given by the following equation:

wherein C is_str,C_feCoefficient of stray losses and iron losses, omega, respectively_eIs the electrical angular velocity, R_sAnd R_fResistance for armature winding and field winding, L_dAnd L_qD, q-axis inductance, T, of the armature winding_eIs an electromagnetic torque, p_oeFor number of pole pairs, psi, of the motor_fIs a permanent magnet flux linkage, K_fAdjusting the magnetic coefficient of the excitation winding;

solving the above equation for_dThe following conclusions can be drawn:

proved by qualitative mathematics, under a certain steady-state working condition, P_lossIs about_dI.e. there exists a certain optimum d-axis flux linkage value psi_{d_opt}Let P_lossAnd minimum.

4. The method of claim 2, wherein the efficiency optimization strategy based on disturbance observation is as follows:

by observing and comparing the current sampling period with the previous sampling period P in real time_lossAnd psi_dDetermines the next sampling period psi_dTo achieve automatic finding of the optimal flux linkage solution psi_{d_opt}。

5. The method according to claim 2, wherein the flux linkage step size in combination with the per-unit input power gradient is selected according to the following principle:

as shown in claims 3 and 4, P is theoretically proved_lossIs about_dIs takingOn-line optimizing direction optimal solution psi_{d_opt}In the process of approximation, P_lossAbout psi_dThe gradient of the input power is gradually reduced to 0, and by combining the characteristic, the variable-step optimization can be realized by taking the input power gradient as a step coefficient in a per unit way in the first few periods of optimization, so as to accelerate the convergence speed and simultaneously realize the improvement of the efficiency, wherein the per unit input power gradient formula is as follows:

。

6. the method of claim 2, wherein the initial step size based on the weak magnetic margin is selected according to the following principle:

the invention defines a variable delta named weak magnetic margin according to the electromagnetic characteristic of a research object_v：

Wherein k is the modulation coefficient of DS-SVPWM, and the value is 0.525; u shape_dcIs a dc bus voltage;

further combining with simulation analysis, introducing a coefficient alpha, and establishing a relation between the weak magnetic margin and the flux linkage step length as follows:

binding delta_vAnd alpha, i.e. the initial step size delta psi of the optimization algorithm can be determined_{d_init}＝0.618×δ_v/α。

7. The method of claim 2, wherein the current distribution mode principle based on the motor torque flux linkage model is as follows:

as described in claims 3, 4, 5, and 6, after obtaining the size of the d-axis flux linkage step in each optimization cycle, the invention proposes an assignment method combining the weak saliency characteristics of the study object in terms of how to implement the assignment method of the d-axis current to the inner and outer stators, and the specific principle is as follows:

consider the flux linkage torque equation for an electric machine:

combining the theoretical proof process of claim 3, considering the weak saliency of the motor and the stability of the motor during the steady-state optimization process, to optimize i during the process_qo1Invariably is an additional condition, and a distribution formula of the d-axis current of the inner stator and the outer stator is obtained:

。

8. the method of claim 2, wherein the modified Rosenbrock algorithm combined with weak magnetic margin is based on the following principle:

based on the idea of the traditional Rosenbrock algorithm, the improved Rosenbrock algorithm provided by the invention brings the defined weak magnetic margin index into the switching condition, further ensures the stability and convergence of the optimizing process, and the improved Rosenbrock algorithm is as follows:

。

Technical Field

The invention belongs to the technical field of motor control, and particularly relates to a hybrid excitation motor weak magnetic area steady-state efficiency optimizing control method.

Background

The hybrid excitation motor is used as a double excitation source type motor and is an extension relative to a single excitation source motor. The novel alternating pole hybrid exciter motor solves the problem that a permanent magnetic field of a traditional permanent magnet synchronous motor is difficult to adjust, has high power density, high torque density and excellent speed regulation performance of the permanent magnet motor, and is particularly suitable for electric automobile application occasions. However, due to the additionally introduced excitation winding and the structural particularity of the motor, the efficiency of the hybrid exciter motor which works in a weak magnetic area is low. Therefore, how to improve the operating efficiency of the weak magnetic region of the hybrid excitation motor by a control means becomes a hot point of domestic and foreign research.

The efficiency optimization control method of the hybrid excitation motor at the present stage can be divided into offline and online efficiency optimization technologies. The off-line efficiency optimization technology realizes the optimal efficiency by solving the optimal efficiency working point under different working conditions and changing the given value in real time, and the working difficulty of the algorithm mainly lies in the construction of a loss model, the effect depends on the accuracy of the model and parameters seriously, the motor loss model is difficult to accurately establish aiming at the parameter disturbance caused by frequent working condition change under the application background of the electric automobile, and the optimization effect is poor. The online efficiency optimization technology changes the working point of the motor in a tentative mode by following a certain specific rule (interval division, step length optimization searching and the like) in each search period, so that the motor is continuously and iteratively transited to the working point with optimal efficiency. The online efficiency optimization technology obtains stronger robustness by sacrificing the convergence speed, but the longer convergence time and the larger current and torque fluctuation caused by the random search process make the online efficiency optimization technology difficult to be applied to the electric automobile application occasions with frequent acceleration and deceleration.

Disclosure of Invention

In view of the above-mentioned shortcomings in the prior art, the present invention provides a method for optimizing the steady-state efficiency of the weak magnetic area of a hybrid excitation motor.

The purpose of the invention can be realized by the following technical scheme:

the method for optimizing the steady-state efficiency of the weak magnetic area of the hybrid excitation motor comprises the following steps:

firstly, a loss model and characteristics of novel alternate pole mixed excitation of a study object are deduced:

novel alternating pole hybrid excitation motor relates to d and q axis magnetic linkage psi_d、ψ_qThe controllable losses (copper, iron, stray losses) are given by the following formula:

wherein C is_str,C_feCoefficient of stray losses and iron losses, omega, respectively_eIs the electrical angular velocity, R_sAnd R_fResistance for armature winding and field winding, L_dAnd L_qD, q-axis inductance, T, of the armature winding_eIs an electromagnetic torque, p_oeFor number of pole pairs, psi, of the motor_fIs a permanent magnet flux linkage, K_fAnd adjusting the magnetic coefficient of the excitation winding.

Further, the above equation is related to_dThe second derivative of (A), P is found_lossIs a concave function about the d-axis flux linkage, i.e. for a certain operating condition, there is a certain optimal d-axis flux linkage psi_{d_opt}The value minimizes the controllable losses.

Further, based on the above inference, a hybrid excitation motor driving control system is built based on a double space vector control technology (DS-SVPWM), and the sampled current and the command voltage output by the PI regulator are used not only to generate a switching signal of the driving system but also to calculate instantaneous power, flux weakening margin and d-axis flux linkage of the driving system in real time.

Further, based on the above inference and driving system basis, the forward direction of the d-axis flux linkage step length in the next optimization period is determined according to the theoretical principle of disturbance observation. The theoretical basis of disturbance observation is as follows:

the instantaneous power P of the current in the current period and the command voltage are sampled and calculated according to an electromagnetic power formula_e(k) To replace P_loss(k) As an observed quantity, the d-axis flux linkage psi is simultaneously calculated_d(k) As a control quantity, a value P corresponding to the previous cycle follows_e(k-1)，ψ_d(k-1) comparison, based on P_loss＝f(ψ_d) The function characteristic is obtained, and the d-axis flux linkage of the next period is obtainedThe forward direction of (3). The electromagnetic power calculation formula is as follows:

P_e＝2.5(u_do1i_do1+u_qo1i_qo1)+1.5(u_dii_di+u_qii_qi)

wherein u is_do1、u_qo1、u_di、u_qiNamely the inner and outer stator voltage command value v_dq(k)、v_dqi(k)。

Further, with respect to P_e(k) In place of P_loss(k) The theoretical basis for the observed quantity is as follows:

in steady state conditions, the output power is constant, P_e≈P_in＝P_loss+P_outThus can use P_eCharacterization of P in the optimization Process_lossThe variation of (2).

Further, based on the theoretical derivation, the driving system basis and the efficiency optimization principle, the specific optimization process comprises the following steps:

s1, after the system to be driven receives the given optimizing start signal, in the first period of the algorithm, the iterative direction of the d-axis flux linkage for improving the motor efficiency is found out tentatively in the mode of the given step length, and the initial calculation power is stored as a basic value P_basis. Meanwhile, the variable quantity delta of the weak magnetic margin in the fixed-step tentative optimization process_vAnd solving the value of alpha, and establishing a relation between the weak magnetic margin and the d-axis flux linkage.

S2, in the second period of the optimization, based on the calculated value delta of the weak magnetic margin_vAnd alpha, determining a variable step length optimization initial value so as to improve the convergence speed and ensure the stability of the motor in the optimization process.

S3 starting from the third period of the optimization, with delta phi_{d_init}Determining the flux linkage step length of each optimization period as initial step length and per unit input power gradient value as step length coefficient, and realizing rapid convergence, i.e.Δψ_{d_vs}＝Grad_Pe*×Δψ_{d_init}。

S4 Grad near the efficiency optimum_PAutomatically decreasing to 0 to achieve automatic convergence of the step size, in fact, when the step size is small, P_eThe power increase and decrease caused by the change of the working point of the motor can not be effectively reflected. In order to enhance the robustness of the algorithm, when the step length is smaller than a certain threshold value, the algorithm is automatically switched to a Rosenbrock strategy, and the convergence is ensured while the convergence speed is ensured.

Further, the weak magnetic margin definition in S1 is a completely new variable defined in consideration of the possibility of the weak magnetic runaway existing in the fixed-step optimization process of the hybrid excitation motor, in combination with the electromagnetic characteristics thereof, and the formula thereof is as follows:

wherein k is the modulation coefficient of DS-SVPWM, and the value is 0.525; u shape_dcIs the dc bus voltage.

Besides, based on simulation analysis of fixed-step optimization mode, delta is found_vAnd flux linkage step lengthIs chosen approximately linearly. Therefore, a coefficient α is further introduced to link the weak magnetic margin with the flux linkage step size, as follows:

further, the selection formula of the initial step value in S2 is as follows:

Δψ_{d_init}＝0.618×δ_v/α

the coefficient of 0.618 is selected based on the unknownness of the efficiency optimum point before optimization and the consideration of convergence speed and algorithm stability.

Further, the definition of the per-unit-based input power gradient coefficient in S3 is mainly derived from the above description about P_lossDerivation of the properties from the initial point to the optimal d-axis flux linkage psi_{d_opt}During the transition, the input power gradient Grad_PIs gradually decreased to 0. The variable is used as a coefficient to realize the function of changing the step length.

Further, the modified rosenblock algorithm in S4 is specifically expressed as follows:

wherein, the traditional Rosenbrock idea is to observe the quantity P_lossWhen the change direction of the control quantity delta psi is reversed, the direction of the control quantity delta psi in the next period is automatically changed, the amplitude of the control quantity delta psi is reduced to be 1/2, and the automatic convergence of the algorithm is realized. The core idea of the modified rosenblock algorithm can be expressed as follows: when the flux linkage step length is smaller than a certain threshold value, the optimization strategy is switched to a Rosenbrock algorithm, the optimization is continued in a fixed step length mode, and when the observed quantity is increased or the calculated value of the weak magnetic margin is smaller than 0, the controlled quantity in the next periodThe step size is automatically decreased and reversed until convergence.

Furthermore, how to implement the variable quantity of the d-axis flux linkage in each step length in the current distribution mode is obtained. Weak saliency and P based on study object_loss＝f(ψ_d) The invention introduces an additional condition i_qo1Const. The specific distribution of the current can be obtained by combining the motor torque and the flux linkage formulaAnd (4) a formula. The specific formula is as follows:

the invention has the beneficial effects that:

1. the method is based on the disturbance observation idea, realizes efficiency optimization in a variable step length mode, and has strong algorithm robustness and low dependence on parameter accuracy.

2. The invention adopts the observed quantity to calculate the instantaneous power P_eIn place of P_lossThe calculated amount is small, a power measuring device does not need to be additionally introduced, the cost is low, other losses such as iron loss and stray loss of the motor except copper loss can be considered, and the operation efficiency of the motor is optimized from the perspective of a system.

3. Based on motor loss characteristic analysis, the method adopts per-unit input power gradient as a step coefficient to realize variable step search, and ensures the convergence of the algorithm while considering the search speed.

4. The invention defines a weak magnetic margin variable based on the electromagnetic characteristic change condition in the weak magnetic area optimizing process of the hybrid excitation motor, and solves the problem of difficulty in determining the initial step length in a variable step length algorithm based on the variable. In order to ensure the safety in the optimizing process, the variable is also applied to a logic judgment module and a convergence module of the optimizing algorithm, so that the possibility of the out-of-control flux weakening is avoided.

5. The weak magnetic region efficiency optimizing method based on the input power gradient method has the advantages of strong robustness, high safety, high convergence speed, simple realization while improving the system efficiency, and easy popularization and application.

Drawings

In order to more clearly illustrate the embodiments or technical solutions in the prior art of the present invention, the drawings used in the description of the embodiments or prior art will be briefly described below, and it is obvious for those skilled in the art to obtain other drawings without creative efforts;

FIG. 1 is a functional block diagram of a hybrid excitation motor efficiency optimization control system;

FIG. 2 is a schematic diagram of an efficiency optimization strategy based on disturbance observation;

FIG. 3 is a schematic diagram of a logic decision module implementing an efficiency optimization strategy;

FIG. 4 is a flow chart for implementing variable step size optimization based on a combination of an input power gradient method and weak magnetic margin.

Detailed Description

The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.

The present invention will be further described with reference to the accompanying drawings.

First, a loss model and characteristics of a novel alternate-pole hybrid excitation of an object to be studied are derived.

Based on a general motor system loss analysis method and combined with the structural characteristics of a research object, the novel hybrid excitation motor has a d-axis and q-axis flux linkage psi_d、ψ_qThe controllable losses (copper, iron, stray losses) are given by the following formula:

further, the above equation is related to_dThe second derivative of (A), P is found_lossIs a concave function about the d-axis flux linkage as described by:

in conjunction with fig. 2, the following two conclusions can be drawn:

1. aiming at a certain operation condition, a certain optimal d-axis flux linkage psi exists_{d_opt}The value corresponds to the solution of equation (3) such that the controllable loss is minimal.

2. The closer the current working point of the motor is to the optimal point psi_{d_opt}，P_lossAbout psi_dModulus of gradient | Grad_PThe smaller the | is, when the motor operates at the optimal operating point psi_{d_opt}When the corresponding gradient value is 0, i.e. Grad in FIG. 2_P3Is shown as 0.

Based on the above conclusion 1, consider C_str,C_feThe difficulty of obtaining an accurate value and the higher requirement of an analytic algorithm on the accuracy of parameters adopt a disturbance observation on-line optimization mode under a certain steady-state working condition to enable the working point of the motor to transit to an efficiency optimal point. Based on the above conclusion 2, the gradient index Grad_PAnd weak magnetic margin delta defined based on electromagnetic characteristics of hybrid excitation motor in optimization process_vThe method is introduced into an online optimization algorithm, so that the stability of the algorithm is ensured while the convergence speed of the algorithm is improved.

Then, a driving control system of a research object is constructed based on the theoretical derivation, and a schematic block diagram of the efficiency optimizing control system of the hybrid excitation motor is given in fig. 1. The construction of the driving system is based on a vector control principle, wherein an armature winding is controlled by five-phase double-space adjacent four vectors SVPWM, and an excitation winding is controlled by three-phase SVPWM.

Before the optimization is started, phase current, torque and rotating speed signals of an inner stator and an outer stator of a motor at the current moment (k moment, which becomes k-1 moment after the optimization is started) are acquired in real time through samplingObtaining the current (i) of the motor under a synchronous rotating coordinate system through five-phase coordinate transformation (Clarke transformation and Park transformation)_dq(k)、i_dqi(k) Then the command voltage (v) is obtained through a PI regulator by a rotating speed ring and a given current value command_dq(k)、v_dqi(k) And into the SVPWM module to generate motor drive signals to operate the motor at a given speed.

While performing the above operations, the control system will also sample the resulting current (i)_dq(k)、i_dqi(k) And a command voltage (v)_dq(k)、v_dqi(k) Via power, flux linkage and weak magnetic margin calculation modules to get psi_d(k)、P_e(k) And delta_vThe calculation formula of the real-time values of (1) is as follows:

ψ_d＝L_di_do1+ψ_fo+K_fi_di (4)

P_e＝2.5(u_do1i_do1+u_qo1i_qo1)+1.5(u_dii_di+u_qii_qi) (5)

wherein the weak magnetic margin is defined based on the following consideration: due to the change of the electromagnetic characteristics of the hybrid excitation motor in the weak magnetic region, under certain working conditions, before optimization reaches an optimal solution, the working point of the motor may work outside a voltage limit circle, so that the weak magnetic field is out of control. Besides, the existing variable-step online optimization methods have the problem that the initial step is difficult to determine. Aiming at the problems, the invention defines the formula (6) as weak magnetic margin by combining the electromagnetic characteristics of the research object and applies the weak magnetic margin as the operation index of the motor to the online optimization algorithm.

After the optimization starts, the algorithm updates its step size in the manner shown in fig. 4, and the specific process is described as follows:

s1: during the first cycle, the control system will accomplish the following three tasks:

1. by fixed step length, heuristicSex finding P_lossThe direction of the decrease is illustrated in fig. 2, which specifically describes the principle:

if the direction of the current periodic probe is the observed quantity P_loss(in actual systems with P_eInstead) is reduced, i.e. Δ P_loss(ΔP_e)<0, keeping the direction unchanged in the next period, and only changing the optimization step length; if the direction is probed to make the observed quantity P_lossIncrease, i.e. Δ P_loss(ΔP_e)>0, then the flux linkage iteration direction is automatically reversed in the next period. In view of the unknown relative positions of the working point and the optimal point of the motor before optimization and the uncertainty of the trial direction, four possibilities are given in fig. 3, and the logic judgment module based on fig. 3 can realize the automatic identification of the direction.

2. Storing the calculated power P at this time_eAs a power base value P_basis。

3. Differencing to determine δ before and after fixed step heuristic optimization_vThe variation, the corresponding alpha value is calculated to establish delta_vAnd Δ ψ d, the specific relationship being as in formula (7):

wherein the definition of α is derived from Δ δ_vA simulation conclusion in a certain linear relation with delta psi d and the aim of establishing the relation between the weak magnetic margin and the flux linkage step length.

S2: during the second cycle, the control system will perform the following two tasks:

1. based on delta_vCalculating initial step length of flux linkage, and making its working point move toward optimum efficiency point psi in iteration direction determined in previous period_{d_opt}Transition, the calculation formula of the step length is as follows:

Δψ_{d_init}＝0.618×δ_v/α (8)

the coefficient of 0.618 is selected based on the unknownness of the efficiency optimum point before optimization and the consideration of convergence speed and algorithm stability.

2. Storing initial step size | delta psi of flux linkage_{d_init}And | is used as a flux linkage base value.

S3: after the third cycle, the control system will complete the following tasks:

updating the self optimizing step length according to the following formula to ensure convergence and simultaneously realize quick optimization:

Δψ_{d_vs}＝Grad_Pe*×Δψ_{d_init} (10)

notably, the inference drawn from the schematic diagram of fig. 2 is: grad_PWill be separated from the operating point by psi_{d_opt}The distance is reduced to 0 gradually so that at Δ ψ_{d_init}Under the premise of no change, the algorithm based on the formula (10) realizes the automatic update of the step size.

S4: in the last cycle, close to the optimal point, the control system will perform the following tasks:

after the switching condition of the Rosenbrock algorithm is met, the driving system can automatically switch, judge the optimizing direction and update the step length of the driving system according to the principle of the Rosenbrock algorithm until the optimal efficiency working point is converged, the principle of the Rosenbrock algorithm is described as the following formula, and the specific core idea is described in relevant parts of the invention content.

After the flux linkage step length of each optimization cycle is obtained based on the flowchart of fig. 4, the algorithm will follow a certain rule (based on the weak saliency and P of the study object)_loss＝f(ψ_d) An additional condition i introduced by the function characteristic derivation process_qo1Const) is converted into a current instruction, a motor driving signal is generated through a PI controller and an SVPWM algorithm module, so that the working point of the motor is continuously transited to the optimal point of efficiency, and a specific distribution formula is as follows:

in the description of the present invention, the above is merely an illustration of one embodiment of the present invention, and it should be noted that the above embodiment does not limit the present invention, and various changes and modifications made by workers within the scope of the technical idea of the present invention fall within the protection scope of the present invention.