Control parameter determination method, electronic equipment and storage medium

Abstract

The invention relates to a control parameter determination method, electronic equipment and a storage medium, and the method comprises the steps: obtaining at least one performance index required by an optical image stabilization system, determining an associated candidate control parameter according to the performance index, determining a candidate control parameter value combination according to the candidate control parameter, and determining a control parameter value combination according to the candidate control parameter value combination; obtaining a performance evaluation function associated with a performance index of the optical image stabilization system, and determining a control parameter value associated with the performance index in the control element based on the candidate control parameter value combination and the performance evaluation function; according to the method, a traditional manual trial and error process is replaced by automatically generating a candidate control parameter value combination, inconsistency caused by subjective judgment of an engineer is eliminated through an established quantified performance evaluation function, and the performance evaluation accuracy is improved through the synergistic effect of the method for automatically generating the candidate control parameter value combination and the quantified performance evaluation function. And the determination efficiency of the control parameters of the control element is obviously improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of optical image stabilization, and particularly relates to a control parameter determination method for optical image stabilization, an electronic device and a storage medium. BACKGROUND

[0002] In an optical image stabilization (OIS) system, hereinafter referred to as an optical stabilization system, the setting of control parameters of a control component (such as a controller and / or a filter, etc.) has a decisive influence on the performance of the system.

[0003] The current control parameter setting of the OIS system mainly has the following technical problems: first, most manufacturers still rely on experienced engineers to set the control parameters through manual debugging, and this manual debugging process not only consumes time and effort, but also is difficult to realize standardization, resulting in significant differences in parameters debugged by different engineers; second, the industry lacks a systematic parameter optimization process, making it difficult to ensure the performance consistency between different product batches and affecting the product quality stability.

[0004] The root cause of these problems lies in the fact that the setting process of the control parameters excessively relies on the personal experience of engineers and lacks a scientific and unified optimization method.

[0005] In view of the above problems, the prior art needs to be improved. SUMMARY

[0006] The embodiments of the present application provide a control parameter determination method for optical image stabilization, an electronic device and a storage medium, which have the advantages of improving parameter optimization efficiency, ensuring parameter setting consistency and realizing multi-target balanced optimization, so as to at least partially solve the above technical problems.

[0007] In order to achieve the above purpose, the embodiments of the present application provide a control parameter determination method for determining at least one control parameter value of a control component in optical image stabilization, comprising:

[0008] Obtaining at least one performance index required by the optical stabilization system, and determining an associated candidate control parameter according to the performance index;

[0009] Determining a candidate control parameter value combination according to the candidate control parameter;

[0010] Obtaining a performance evaluation function associated with the performance index of the optical stabilization system;

[0011] Determining a control parameter value associated with the performance index in the control component based on the candidate control parameter value combination and the performance evaluation function.

[0012] The embodiment of the present application further provides an electronic device, comprising a memory having a computer program stored thereon, and a processor configured to execute the computer program in the memory to implement the control parameter determination method.

[0013] The embodiment of the present application further provides a computer readable storage medium having a computer program stored thereon, and the computer program is executed by a processor to implement the control parameter determination method.

[0014] The embodiment of the present application provides a control parameter determination method, an electronic device and a storage medium. The method generates candidate control parameter value combinations automatically, replaces the traditional manual trial and error process, systematically covers the parameter space, avoids the problem that potential optimal solutions are missed, eliminates the inconsistency caused by subjective judgment of engineers through the established quantitative performance evaluation function, and significantly improves the determination efficiency of the control parameters of the control element through the synergistic effect of the method of automatically generating candidate control parameter value combinations and the quantitative performance evaluation function.

[0015] Other features and advantages of the present application will be described in detail in the following specific embodiments. BRIEF DESCRIPTION OF DRAWINGS

[0016] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the following will briefly introduce the drawings needed to be used in the embodiment description. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can also be obtained by those skilled in the art without any creative effort on the basis of these drawings.

[0017] In order to more completely understand the present application and its beneficial effects, the following will be described in conjunction with the drawings, wherein the same reference numerals in the following description represent the same parts.

[0018] Figure 1 is a flowchart of the control parameter determination method provided by the embodiment of the present application;

[0019] Figure 2 is a sub-flowchart of the control parameter determination method provided by the embodiment of the present application;

[0020] Figure 3 is a structure diagram of the control parameter determination device provided by the embodiment of the present application;

[0021] Figure 4 is a schematic diagram of the electronic device provided by the embodiment of the present application. DETAILED DESCRIPTION

[0022] With reference to the drawings, the technical solutions in the embodiments of the present application will be clearly and completely described below. Obviously, the described embodiments are only a part of the embodiments of the present application, but not all the embodiments of the present application. Based on the embodiments in the present application, all other embodiments obtained by a person of ordinary skill in the art without creative effort are within the protection scope of the present application.

[0023] Before the embodiments of the present application are described in detail, the related art is further described and analyzed.

[0024] Optical image stabilization technology has become a key component for improving imaging quality in the field of consumer electronics, especially in the camera of a smart phone. The performance of an optical image stabilization system depends largely on the accurate setting of control parameters (such as parameters of a controller, and / or parameters of a filter, etc.). In the related art, the setting of the control parameters is usually obtained by manual debugging by an engineer.

[0025] In the process of debugging the control parameters of a traditional optical image stabilization system, an engineer needs to manually adjust a set of parameters such as proportional gain, integral gain, and differential gain of a controller, and / or parameters such as cutoff frequency and gain parameters of a filter. This process relies on the experience accumulation and trial-and-error verification of a debugging personnel such as an engineer. Due to the high dimensionality of the parameter space and the nonlinear coupling relationship between the parameters, manual debugging is difficult to systematically cover key parameter combinations, resulting in low parameter optimization efficiency. At the same time, different engineers have subjective differences in the evaluation criteria of performance indicators, resulting in a lack of repeatability of parameter setting results. For example, the same model of product in different production batches has inconsistent technical indicators such as a fluctuation range of dynamic response overshoot exceeding a set range, and / or a discrete degree of step signal steady-state error exceeding a preset error.

[0026] If the above problems are not solved, the debugging period of the optical image stabilization system will increase exponentially with the increase of the parameter dimension, which seriously restricts the development progress of new products. The randomness of parameter setting will cause the key performance indicators of mass production modules to exceed the design tolerance, causing problems such as shift of system resonance frequency, and / or large fluctuation of closed-loop bandwidth. At the same time, the parameter combinations selected by different engineers according to different evaluation criteria of performance indicators are difficult to cover the optimal parameter region, which may cause the degradation of the signal-to-noise ratio of the output signal of the image sensor. All of these will directly affect the imaging quality and market competitiveness of the optical image stabilization system related products.

[0027] To this end, the application provides a control parameter determination method, a control parameter determination device, an electronic device, a computer readable storage medium and a computer program product, which mainly improve in the following two aspects: first, an automatic parameter generation strategy is used to replace manual selection of parameters; second, a multi-dimensional performance evaluation system is established to convert the subjective experience of engineers into quantifiable scoring standards. The application optimizes the parameter generation and the evaluation mechanism of the performance evaluation in a coordinated manner to form a closed-loop optimization process.

[0028] A control parameter determination method, a control parameter determination device, an electronic device, a computer readable storage medium and a computer program product provided by the application are described in detail below. The control parameter determination method of the application is applied to an electronic device, and the control parameter determination device of the application can be integrated in an electronic device.

[0029] As shown in Figure 1 The application embodiment provides a control parameter determination method for determining at least one control parameter value of a control member in optical image stabilization. Specifically, the control parameter determination method provided by the application embodiment includes the following steps.

[0030] S101, at least one performance index required by an optical image stabilization system is obtained, and a candidate control parameter associated with the performance index is determined.

[0031] First, at least one performance index required by the optical image stabilization is obtained. For example, the design requirements of the optical image stabilization system are obtained, and the performance index for measuring the control member of the optical image stabilization system is set according to the design requirements. Specifically, the design requirements of the optical image stabilization system are determined according to the market positioning, application scenario or user experience target of the product corresponding to the function of the optical image stabilization system. For example, taking a smart phone as an example, the camera of the mainstream mobile phone mainly deals with daily shaking, while the camera of the motion camera needs to deal with more severe shaking and support video super anti-shake function, etc. Different performance indexes are set according to different requirements of the shaking function.

[0032] The performance index for measuring the control member of the optical image stabilization system has one or more, including but not limited to one or more of the model parameters of the controlled object such as the motor, and / or one or more of the static parameters, and / or one or more of the dynamic parameters, and / or one or more of the control loop effects, and / or one or more of the image effects, etc. The performance index in the application embodiment covers the whole link of hardware characteristics, control effect and final imaging quality, ensuring that the parameter optimization method of the application embodiment meets the performance requirements in multiple dimensions.

[0033] The motor model parameters include, but are not limited to, performance indicators such as a resonance frequency F0 of the controlled object (e.g., a motor) and a resistance value FP0. The static parameters describe the characteristics of the optical image stabilization system in a slow change or static state, including but not limited to performance indicators such as a rated stroke of the controlled object (e.g., a motor) that can be stably moved, linearity, sensitivity, hysteresis, and the like. The dynamic parameters are used to describe the response capability of the optical image stabilization system to a rapidly changing input signal, including but not limited to performance indicators such as a steady-state response time of a step signal, an overshoot, a steady-state error, a sinusoidal wave following delay / accuracy, and the like. The control loop effects include, but are not limited to, performance indicators such as OpenLoop Frequency Response Analysis (OpenLoop FRA), ClosedLoop Frequency Response Analysis (ClosedLoop FRA), Phase Margin (PM), Gain Margin (GM), and the like. The image effects include, but are not limited to, performance indicators such as Modulation Transfer Function (MTF), Suppression Ratio (SR), and the like. It should be noted that the performance indicators listed here are only for illustrative purposes and do not constitute a limitation on related performance indicators.

[0034] After the performance indicators of the optical image stabilization system are determined according to the design requirements, the associated candidate control parameters in the control components are determined according to the performance indicators. The control components include the controller and / or other control components such as filters, as described above. In the embodiments of the present application, the control components include the controller and the filter as an example for description.

[0035] The control parameters of the controller are different from the control parameters of the filter. For example, the control parameters of the controller include a proportional gain (denoted by K P ), an integral gain (denoted by K I ), a differential gain (denoted by K D ), and the like, and the control parameters of the filter include a cutoff frequency (denoted by f c ) of the filter, a gain, and the like.

[0036] The candidate control parameters associated with the performance indicators are determined from the control parameters of the control components according to the performance indicators, so that the candidate control parameters are determined according to the performance indicators, so that the candidate control parameters correspond to the performance indicators, and the finally determined control parameter values are more accurate.

[0037] S102, determine a candidate control parameter value combination according to the candidate control parameters.

[0038] The candidate control parameter value combination refers to a set of possible candidate control parameter values screened according to the candidate control parameters, that is, a set of adjustable parameter values of the performance indicators affecting the requirements of the optical image stabilization system.

[0039] For each candidate control parameter, first, a search space / parameter space of the parameter value corresponding to each candidate control parameter is defined, that is, a candidate parameter value interval, for example, the candidate parameter value interval of the proportional gain of the controller is K P :[0.5,2.5], and then the candidate control parameter value combination is determined according to the candidate parameter value interval corresponding to each candidate control parameter. A corresponding search method or optimization method can be used to determine from the parameter space of the corresponding candidate control parameter, for example, the candidate control parameter value combination is determined by equally spaced sampling of each candidate parameter value interval.

[0040] Each candidate control parameter value combination determined includes the parameter values of the candidate control parameters, and the parameter values of the candidate control parameters are all in the corresponding candidate parameter value interval, so that multiple candidate control parameter value combinations can be obtained.

[0041] The candidate control parameter value combination realizes systematic coverage of each candidate control parameter in the corresponding parameter space, avoiding the randomness of manual debugging.

[0042] In S103, a performance evaluation function associated with the performance indicators of the optical image stabilization system is obtained.

[0043] The performance evaluation function is associated with multiple performance indicators of the optical image stabilization system, and the performance evaluation function refers to a mathematical model for converting multiple performance indicators into quantitative scores. Specifically, the scores of one or more of the motor model parameters, static parameters, dynamic parameters, control loop effects, and image effects can be integrated into a single numerical value by using a weighted summation method, for example, the performance evaluation function can quantitatively score multiple performance indicators and combine them by weighting to form a total evaluation standard for the candidate control parameter value combination. The performance evaluation function provides a comparable evaluation standard for different parameter combinations by establishing an index weight distribution mechanism and a segmented scoring rule.

[0044] The design requirements of the optical image stabilization system are different, and the corresponding performance indicators are different, so that the performance evaluation function is also different. The performance evaluation function can be determined in advance according to the performance indicators, or can be determined in real time after the performance indicators are determined.

[0045] The performance evaluation function is associated with multiple performance indicators, and the embodiments of the present application quantitatively convert multiple performance indicators into a unified standard, so that different candidate control parameter value combinations can be objectively compared, avoiding errors caused by engineers according to experience.

[0046] S104, determine the control parameter value associated with the performance index of the optical image stabilization system in the control element based on the candidate control parameter value combination and the performance evaluation function.

[0047] After the candidate control parameter value combination and the performance evaluation function are determined, the performance of each candidate control parameter value combination is evaluated, each candidate control parameter value combination is substituted into the system model or the actual system, the verification items are verified, the performance indexes are tested, the values of the performance indexes are input into the associated performance evaluation function, the score results (which can also be referred to as performance evaluation values) of the performance evaluation function are obtained, and finally the candidate control parameter value combination corresponding to the optimal score result is selected, and the optimal candidate control parameter value combination is used as the control parameter value of the final control element.

[0048] Through the above scheme, the embodiment of the present application realizes the automatic optimization of the control parameter of the control element of the optical image stabilization system. The embodiment of the present application automatically generates the candidate control parameter value combination, replaces the traditional manual trial-and-error process, systematically covers the parameter space, and avoids the problem that the potential optimal solution is missed; through the established quantitative performance evaluation function, the inconsistency caused by the subjective judgment of the engineer is eliminated. Through the synergistic effect of the method of automatically generating the candidate control parameter value combination and the quantitative performance evaluation function, the debugging efficiency of the control parameter of the control element is significantly improved, and the debugging process originally measured in days is shortened to hours; at the same time, the standardized optimization process ensures the performance consistency between different batches of products, and effectively solves the performance fluctuation problem caused by the randomness of the parameter setting in the traditional method.

[0049] Among them, if only the method of automatically generating the candidate control parameter value combination is used, although the search range can be expanded, the optimization direction deviation problem caused by the inconsistent evaluation standard of the performance evaluation cannot be solved; if only the performance evaluation function is established without changing the parameter generation method, the randomness of the manual parameter selection is still limited. Therefore, in the embodiment of the present application, the method of automatically generating the candidate control parameter value combination and the quantitative performance evaluation function are synergistically used to improve the debugging efficiency and consistency of the control parameter.

[0050] As described above, the control element in the embodiments of the present application includes a controller, however, there are many types of controllers, and different controller types correspond to different control parameters. The control parameters of the controller need to be determined according to the type of the controller. For example, if the type of the controller is a common PID controller, it corresponds to three parameters of proportional gain, integral gain and derivative gain; if the type of the controller is a feedforward PID controller, it corresponds to four parameters of feedforward parameter, proportional gain, integral gain and derivative gain. Different filter types also correspond to different control parameters, such as high-pass filter, low-pass filter and band-pass filter, and the control parameters of the filters are also different. The control parameters of the filters also need to be determined according to the type of the filter.

[0051] In an embodiment, the step of determining the candidate control parameter associated with the performance index in the control element in the step S101 described above includes: obtaining the type of the control element; and determining the candidate control parameter associated with the performance index from the control parameters matching the type of the control element according to the performance index. Thus, the determined candidate control parameter is clearly related to the performance index and the type of the control element.

[0052] In an optical image stabilization system, there are many types of control elements, and different control element types correspond to different control parameters. The control parameters of the control element need to be determined according to the type of the control element. When the control element includes a controller, different controller types correspond to different control parameters. For example, if the type of the controller is a common PID controller, it corresponds to three parameters of proportional gain, integral gain and derivative gain; if the type of the controller is a feedforward PID controller, it corresponds to four parameters of feedforward parameter, proportional gain, integral gain and derivative gain. Different filter types also correspond to different control parameters, such as high-pass filter, low-pass filter and band-pass filter, and the control parameters of the filters are also different. The control parameters of the filters also need to be determined according to the type of the filter.

[0053] In an embodiment, the step of determining the candidate control parameter associated with the performance index from the control parameters matching the type of the control element according to the performance index includes: determining a first candidate control parameter (part of the candidate control parameter) associated with the performance index from the control parameters matching the type of the control element according to the performance index, and determining a second candidate control parameter (another part of the candidate control parameter) associated with the performance index from the control parameters matching the type of the control element based on the first candidate control parameter, so as to take the first candidate control parameter and the second candidate control parameter as the candidate control parameter. In this way, the other part of the candidate control parameter is determined according to the first candidate control parameter, which takes into account the association between the candidate control parameters, and through the synergistic effect between the candidate control parameters with the association, the accuracy of the finally determined control parameter value is improved.

[0054] In one embodiment, the control unit includes multiple control units, such as a first control unit like a controller and a second control unit like a filter. Correspondingly, the step of determining candidate control parameters associated with the performance index from control parameters matching the type of the control unit according to the performance index includes: determining a third candidate control parameter associated with the performance index from control parameters matching the type of the first control unit according to the performance index; determining a fourth candidate control parameter associated with the performance index from control parameters matching the type of the second control unit based on the third candidate control parameter; and using the third and fourth candidate control parameters as candidate control parameters. Thus, determining the fourth candidate control parameter of the filter based on the third candidate control parameter of the controller takes into account the correlation between the control units. Through the synergistic effect of the candidate control parameters of the controller and the filter, the accuracy of the finally determined control parameter value is improved.

[0055] Since there are multiple different performance indicators in this embodiment, and the candidate control parameters corresponding to each performance indicator are also different, the associated candidate control parameters can be dynamically and automatically determined according to different performance indicators in this embodiment, thereby improving the efficiency and accuracy of determining associated candidate control parameters based on different performance indicators.

[0056] In one embodiment, step S102, namely the step of determining the combination of candidate control parameter values ​​based on the candidate control parameters, includes: outputting a parameter space based on the candidate parameter value range of each candidate control parameter; and outputting the combination of candidate control parameter values ​​based on the parameter selection method and the parameter space.

[0057] The parameter space is constructed by defining the value ranges of each candidate control parameter. These ranges can be continuous numerical ranges or discrete numerical sets, depending on the parameter type and the physically realizable range of each candidate control parameter. For example, the proportional gain is limited to the range [0.5, 2.5] to prevent oscillations in the optical image stabilization system, while the integral gain is limited to the range [0.01, 0.05] to avoid integral saturation. The explicit definition of the candidate parameter value ranges avoids the randomness of the candidate control parameter values ​​while ensuring that the values ​​remain within the physically realizable range.

[0058] For example, for the optical image stabilization module of a certain smartphone, the candidate control parameters include the proportional gain K. P Integral gain K I Differential gain K D and the filter cutoff frequency f c The corresponding candidate parameter value ranges are K. P [0.5, 2.5]、K I [0.01, 0.05]、KD [0.001, 0.01], f c [50Hz, 200Hz]. The candidate parameter value interval of the four parameters constitutes a four-dimensional parameter space.

[0059] After the parameter space is determined, the candidate control parameter value combination is output according to the parameter space based on a parameter selection mode. There are various parameter selection modes, and the parameter selection mode selects a corresponding strategy according to an application scenario, for example, a grid search is adopted to realize global coverage when computing resources are sufficient, and a coarse-to-fine search is adopted to narrow down the search range in stages when resources are limited. The candidate control parameter value combination is determined according to different parameter selection modes, so that a more accurate and more matched candidate control parameter value combination can be obtained.

[0060] In an embodiment, the parameter selection mode includes a search mode, and the step of outputting the candidate control parameter value combination according to the parameter space based on the parameter selection mode includes: outputting the candidate control parameter value combination according to the parameter space based on a search strategy corresponding to the search mode. The candidate control parameter value combination is determined based on the search strategy, avoiding the randomness and omission caused by manual setting of engineers, and improving the determination efficiency of the candidate control parameter value combination.

[0061] The search strategy includes at least one of a grid search and a coarse-to-fine search.

[0062] In the grid search, the candidate parameter value interval of each candidate control parameter is divided into equal intervals. The equal interval division can be realized in a fixed step or a dynamic step manner. Here, a fixed step is taken as an example for illustration. For example, the candidate parameter value interval [0.5, 2.5] of the proportional gain K P is divided into five value points (five discrete values) with a step size of 0.5, the candidate parameter value interval [0.01, 0.05] of the integral gain K I is divided into five value points (five discrete values) with a step size of 0.01, the candidate parameter value interval [0.001, 0.01] of the differential gain K D is divided into ten value points (ten discrete values) with a step size of 0.001, and the candidate parameter value interval [50Hz, 200Hz] of the cutoff frequency f c of the low-pass filter is divided into four value points (four discrete values) with a step size of 50HZ, thereby generating 5x5x10x4=1000 grid combinations, that is, 1000 candidate control parameter value combinations.

[0063] Understandably, in the grid search, the parameter space is discretized into uniformly distributed grid points, and each grid point corresponds to a candidate control parameter value combination. In the grid search, the comprehensiveness of parameter optimization is ensured through comprehensive coverage of the parameter space.

[0064] wherein each candidate control parameter value combination comprises a set of specific K P , K I , K D and f c values. For example, one of the candidate control parameter value combinations can be (K P = 1.5, K I = 0.03, K D = 0.05, f c = 100 Hz). These candidate control parameter value combinations constitute the search space for the subsequent optimization process. A unified verification is performed on all combinations at one time, and after each combination is verified according to the same verification items, the optimal candidate control parameter value combination is selected according to the score results of the associated performance evaluation function.

[0065] In the coarse-to-fine search, the determination of the candidate control parameter value combination is completed in stages. In the first stage, a rough determination is made with a large step size and a first batch of verification items is executed, and candidate control parameter value combinations with high score results are screened out, for example, the top 10% of candidate control parameter value combinations are screened out; in the second stage, the screened candidate control parameter value combinations are locally subdivided and a second batch of verification items is executed, and the optimal candidate control parameter value combination is obtained therefrom.

[0066] wherein the first stage generates primary candidate control parameter value combinations with a large step size, for example, five value points are generated in the candidate parameter value interval [0.5, 2.5] of the proportional gain K P with a step size of 0.5, five value points are generated in the candidate parameter value interval [0.01, 0.05] of the integral gain K I with a step size of 0.01, ten value points are generated in the candidate parameter value interval [0.001, 0.01] of the derivative gain K D with a step size of 0.001, and four value points are generated in the candidate parameter value interval [50 Hz, 200 Hz] of the cutoff frequency f c of the low-pass filter with a step size of 50 Hz, thus generating 5x5x10x4 = 1000 candidate control parameter value combinations. It should be noted that in the coarse-to-fine search, in order to reduce the calculation cost and improve the efficiency, the first stage is allowed to only execute part of the high-weight verification items; correspondingly, part of the high-weight verification items are executed on the 1000 candidate control parameter value combinations to screen out candidate control parameter value combinations with high score results from the 1000 candidate control parameter value combinations.

[0067] After obtaining the candidate control parameter value combinations with higher score results, the selected intervals of each candidate control parameter are refined in the second stage with smaller steps, that is, a local subdivision grid is constructed with the candidate control parameter value combinations with higher score results as the center. For example, the interval [1.3, 1.7] is generated near the proportional gain 1.5 with higher score, the interval [0.25, 0.35] is generated near the integral gain 0.03 with higher score, the interval [0.004, 0.006] is generated near the derivative gain 0.005 with higher score, and the interval [110Hz, 130Hz] is generated near the cutoff frequency 120Hz with higher score. The interval [1.3, 1.7] of the proportional gain is divided with a step size of 0.1, the interval [0.25, 0.35] of the integral gain is divided with a step size of 0.005, the interval [0.004, 0.006] of the derivative gain is divided with a step size of 0.005, and the interval [110Hz, 130Hz] of the cutoff frequency is divided with a step size of 10Hz, thereby obtaining the refined candidate control parameter value combinations. For each refined candidate control parameter value combination, a complete verification item is executed, thereby selecting the candidate control parameter value combination with the optimal score result.

[0068] Understandably, in the coarse-to-fine search, the first stage uses a larger step size to quickly screen potential high-quality parameter regions, locates high-quality parameter regions with fewer verification times, and improves efficiency while ensuring accuracy. The second stage locates the optimal candidate control parameter value combination from the high-quality parameter region. Through dynamic allocation of the number of verification times in two stages, accuracy is ensured while efficiency is improved.

[0069] Both of the above two search strategies (grid search and coarse-to-fine search) need to be combined with the unified verification item in the performance evaluation function, wherein the grid search requires all candidate control parameter value combinations to execute all verification items, and the coarse-to-fine search allows the first stage to execute only part of the high-weight verification items to reduce the calculation cost.

[0070] Whether it is grid search or coarse-to-fine search, these search strategies realize the traversal of the corresponding parameter space, improve the efficiency of parameter optimization, and the efficiency is improved very obviously especially when the parameter space is large. For example, taking the grid search and a four-dimensional parameter space as an example, the number of candidate control parameter value combinations is approximately AxBxNxM, and the random search needs to traverse approximately N^4 possibilities. The combination of parameter selection and parameter space effectively avoids the generation of invalid parameter combinations, and the boundary constraint of the parameter space can prevent the algorithm from iterating to a physically unachievable region, ensuring that each generated candidate control parameter value combination meets the system design requirements.

[0071] In an embodiment, the parameter selection manner comprises an optimization manner, and the step of outputting the candidate control parameter value combination according to the parameter space based on the parameter selection manner comprises: outputting the candidate control parameter value combination according to the parameter space based on the initial parameter value and the optimization strategy corresponding to the optimization manner.

[0072] The optimization strategy can be one of a simplex method or a genetic algorithm.

[0073] The implementation of the simplex method comprises constructing an initial simplex structure in a multi-dimensional parameter space, for example, if it is intended to optimize the proportional gain Kp P , the integral gain K I , and the cutoff frequency f c , three groups of initial parameter points are selected in a three-dimensional parameter space to form an initial simplex triangle, each group of initial parameter points (also referred to as initial parameter values) comprises a proportional gain, an integral gain, and a cutoff frequency, the performance evaluation function score result is obtained by verifying each group of initial parameter points (corresponding to each vertex in the initial simplex triangle) through a verification project, and the score distribution is determined according to the score result. The reflection, expansion, or contraction operation is performed based on the score result, for example, the reflection operation is performed to generate a new vertex, for example, the worst vertex is moved a certain distance along the opposite side to obtain a new vertex, the worst vertex is replaced by the new vertex after reflection, if the score of the new vertex is better than that of the original worst vertex, i.e., the original vertex, the original vertex is replaced and the next iteration is entered; if the score of the new vertex is still the worst, the simplex size is reduced through the contraction operation to adjust the movement of the simplex structure to the parameter region with the optimal score result. The iteration process is repeated until the simplex converges or the maximum number of verifications (for example, 30 times) is reached, and through multiple iterations, the simplex structure gradually converges to the parameter region with the optimal score result.

[0074] The simplex method is used for optimization, and the simplex method can quickly approach the optimal solution by dynamically adjusting the search direction, avoids the search blind area caused by the fixed step length, and improves the efficiency and accuracy of parameter optimization.

[0075] The implementation of the genetic algorithm comprises generating an initial population, for example, 20 groups of initial candidate control parameter value combinations, evaluating each initial candidate control parameter value combination through a verification project to obtain the score result of the performance evaluation function, performing a selection operation according to the score result, and retaining the better candidate control parameter value combinations as parents. The crossover operation is performed on the candidate control parameter value combinations of the parents, for example, single-point crossover or uniform crossover, to generate the candidate control parameter value combinations of the offspring, and the random mutation operation is introduced, for example, a slight disturbance is made to a certain candidate control parameter with a probability of 5% to maintain the population diversity.

[0076] When using a genetic algorithm, an initial population is randomly generated and covers multiple regions of the parameter space, for example, uniformly distributing multiple initial values in a four-dimensional parameter interval, i.e., initial candidate control parameter value combinations. In each iteration, the top 50% of parameter combinations (candidate control parameter value combinations) are retained through the selection operation, and a new parameter combination, i.e., a new candidate control parameter value combination, is generated through the crossover operation, for example, the proportional gain K P An arithmetic mean is taken to form the proportional gain K P of the offspring. The mutation operation introduces a random disturbance with a certain probability, for example, adding a random offset of ±0.005 to the K I value of a certain offspring. After multiple generations of evolution, the performance evaluation function scores of the candidate control parameter value combinations of the population become more and more optimal, and finally converge to the parameter region of the global optimal solution. Through repeated selection, crossover, and mutation, until the convergence condition is met or the maximum number of iterations (for example, 100 generations) is reached.

[0077] The genetic algorithm effectively explores the global optimal solution by simulating the biological evolution process through population iteration and gene operation, improving the probability of finding high-quality candidate control parameter value combinations.

[0078] Both optimization strategies effectively reduce the number of verifications by dynamically adjusting the search direction and combining the verification feedback of the parameter combinations, significantly improving the efficiency and global search ability of parameter optimization. The introduction of the optimization strategy makes the parameter selection process more systematic and automated, reduces the dependence on human experience, and improves the repeatability and consistency of parameter optimization.

[0079] It should be noted that the above-mentioned search methods and optimization methods are only illustrative examples, and other methods can also be used in other embodiments, without limitation.

[0080] In an embodiment, the step S103 described above, i.e., the step of obtaining a performance evaluation function associated with the performance indicators of the optical image stabilization system, includes: outputting an evaluation part in the performance evaluation function based on the evaluation method of each performance indicator and the indicator weight; and outputting a constraint part in the performance evaluation function based on the constraint relationship of each performance indicator. The evaluation part and the constraint part of the performance evaluation function can be displayed in the form of a formula or a function.

[0081] The evaluation method of the performance indicator is used to determine the quantification method for quantifying the performance indicator, and the evaluation method of each performance indicator is different. There are many ways to determine the indicator weight of the performance indicator, which can be dynamically adjusted according to the actual application scenario of the optical image stabilization system, i.e., the indicator weight of the performance indicator is different for different application scenarios.

[0082] The performance evaluation function in the embodiment includes an evaluation part and a constraint part. The evaluation part quantitatively evaluates each performance index, and the constraint part represents the correlation and mutual influence between the performance indexes, so that the performance evaluation function is more accurate, and the control parameter value is further more accurately obtained.

[0083] In an embodiment, the evaluation method of each performance index and the index weight are determined by the following steps: based on the concerned index of the optical image stabilization, the performance indexes and the index weights of the performance indexes are output; and based on the index values and the quantification method of the performance indexes, the evaluation method of each performance index is output.

[0084] Specifically, in the parameter optimization process of the optical image stabilization system, the key performance index / core index affecting the imaging quality is first selected according to the design requirements of the optical image stabilization system, that is, the concerned index. For example, in a closed-loop control system, overshoot, steady-state error and response time are selected as the concerned index, and the concerned index is taken as the corresponding performance index. Through data analysis, it is found that, for example, when the overshoot is more than 10%, the image stabilization time will be prolonged by 30%, and therefore the overshoot is set as the core index.

[0085] The index weight of the performance index can be dynamically adjusted based on the actual application scene of the optical image stabilization system. For example, in a scene requiring fast response, the steady-state response time is given a higher weight; and in a scene requiring stability, the overshoot weight is correspondingly increased. Specifically, for example, the overshoot, the stabilization time and the static error can be selected as the key performance indexes, wherein the overshoot, as a difficult-to-reach performance index, is given a higher weight, and therefore the overshoot, the stabilization time and the static error are respectively given weights of 0.4, 0.3 and 0.3. The way of distributing the index weight is not limited, but the total sum of the index weight needs to meet the normalization requirement.

[0086] The evaluation method of the performance index needs to be combined with the performance threshold (corresponding to the index value) in engineering practice. For example, the overshoot demarcation point can be a preset ratio such as 5%, and the setting is based on the maximum instantaneous offset allowed by the optical assembly, and exceeding the preset ratio will cause image blur. In the quantification process, the performance index can be divided into multiple intervals according to the index value, for example, the overshoot can be divided into overshoot <5%, [5%-10%], [10%-15%] and overshoot >15%, so as to obtain the quantification method of the performance index, each interval corresponds to a different scoring coefficient, so as to obtain the evaluation method of the performance index according to the index value and the quantification method. The evaluation formula of the performance index can be realized by using a piecewise linear function or a piecewise nonlinear function.

[0087] For example, for the overshoot of the performance index, the overshoot is divided into four intervals, each interval corresponds to a different scoring coefficient; for example, the overshoot < 5% is scored 1, the overshoot in [5%-10%] is scored 0.8, the overshoot in [10%-15%] is scored 0.6, and the overshoot > 15% is scored 0; in this way, the evaluation method of the overshoot of the performance index is determined. In specific implementation, a candidate control parameter value combination can be verified by using a verification item to obtain a response curve of the optical image stabilization system, the actual value of the overshoot is extracted from the response curve, and then the performance index score of the overshoot is calculated according to the evaluation method. For example, if the measured overshoot is 7%, the performance index score of the overshoot is 0.8.

[0088] For example, for the steady-state error of the performance index, the steady-state error can also be divided into four intervals, each interval corresponds to a different scoring coefficient, for example, the steady-state error < 0.1° is scored 1, the steady-state error in [0.1°-0.2°] is scored 0.8, the steady-state error in [0.2°-0.3°] is scored 0.6, and the steady-state error > 0.3° is scored 0; thereby obtaining the evaluation method of the steady-state error. In specific embodiments, the actual value of the steady-state error can be obtained by verifying a candidate control parameter value combination using a verification item, and then the performance index score of the steady-state error is calculated according to the evaluation method, for example, the steady-state error is 0.25°, and the performance index score of the steady-state error is 0.6.

[0089] For the performance index, the evaluation method is implemented by using a piecewise function, which improves the sensitivity of the performance index evaluation, makes the performance evaluation more objective and comprehensive, and helps to more accurately identify high-quality candidate control parameter value combinations in the control parameter optimization process, thereby improving the overall performance of the optical image stabilization system. In addition, by converting the actual physical quantity into a standardized score, performance indexes of different dimensions can be comprehensively reflected and compared in a unified framework, so that the parameter optimization process can accurately reflect the design requirements of the optical image stabilization system.

[0090] According to the evaluation method, the performance index score of each performance index can be obtained, and then the evaluation part in the performance evaluation function is determined according to the performance index score and the index weight. For example, the performance index score and the index weight can be weighted to obtain the evaluation part in the performance evaluation function. Specifically, it can be shown as the following formula (1).

[0091] Score=w1S1+w2S2+…+w n S n (1)

[0092] Wherein, S i is the i-th performance index score; w i is the corresponding index weight, and the index weights of the performance indexes satisfy ∑w i= 1.

[0093] In some embodiments of the present application, the performance evaluation function of the plurality of performance indicators, i.e., as shown in equation (1), i.e., the performance evaluation function can only include the evaluation part, in these embodiments, the constraint relationship between the performance indicators is not considered.

[0094] After obtaining the evaluation part of the performance evaluation function, according to the constraint relationship of the performance indicators, the constraint part of the performance evaluation function is determined. In order to enhance the reflection of the performance evaluation function on the performance of the optical image stabilization system, the constraint mechanism of the interaction between the plurality of performance indicators is introduced to construct the constraint part.

[0095] In an embodiment, the step of determining the constraint part of the performance evaluation function according to the constraint relationship of the performance indicators includes: determining a conditional constraint part of the performance evaluation function according to the coupling relationship between the performance indicators, taking the conditional constraint part as the constraint part of the performance evaluation function; or, determining a soft constraint part of the performance evaluation function according to the inhibition relationship between the performance indicators, taking the soft constraint part as the constraint part of the performance evaluation function; or, determining a conditional constraint part of the performance evaluation function according to the coupling relationship between the performance indicators, determining a soft constraint part of the performance evaluation function according to the inhibition relationship between the performance indicators, and determining the constraint part of the performance evaluation function according to the conditional constraint part and the soft constraint part.

[0096] In the following, the determination method of the conditional constraint part and the soft constraint part will be described by way of example. For example, in a typical optical image stabilization system, the following key performance indicators can be set: closed loop frequency response (Closed FRA), overshoot (Overshoot), and open loop frequency response (Open FRA). Among them, the bandwidth of the closed loop frequency response is set as: BandWidth (BW) ≥ 80 Hz, to ensure that the optical image stabilization system has fast response capability; the overshoot is set as: Overshoot ≤ 10%, to avoid excessive system oscillation and ensure image stability; and the phase margin of the open loop frequency response is set as: Phase Margin (PM) ≥ 30°, to ensure that the system has sufficient stability margin.

[0097] In actual optical image stabilization systems, these performance indicators often have coupling, for example, when the bandwidth is increased to improve the response speed, the increase of system gain may cause the increase of Overshoot, and the decrease of PM, which increases the risk of system instability, which can be understood as the coupling relationship between the performance indicators. In order to avoid the search process falling into the local optimal solution caused by such adverse coupling relationship, the conditional constraint part as shown in equation (2) is introduced in the embodiments of the present application.

[0098]

[0099] The condition constraint part is used to force to exclude unstable solutions with lower PM when Overshoot is large, and to strengthen the overall robustness of the optical image stabilization system.

[0100] Further, the application can also introduce a soft constraint part, for example, a soft constraint for insufficient bandwidth, as shown in formula (3).

[0101] g BW = 1 - lambda bw *max(0, 80 - BW) 2 (3)

[0102] The soft constraint part is used to limit the performance evaluation function when the bandwidth is lower than 80Hz, and to apply a penalty / suppression to the candidate control parameter value combination whose performance index does not meet the standard. Wherein, lambda bw is a coefficient, and BW represents the actual value of the bandwidth.

[0103] Finally, the performance evaluation function is determined according to the evaluation part and the constraint part, for example, the evaluation part is multiplied by the constraint part to obtain the performance evaluation function, which can be obtained by using formula (4) as follows.

[0104] Score*g(Overshoot, PM)*g BW (4)

[0105] In the performance evaluation function in the embodiment of the application, the soft and hard constraint parts, such as the condition constraint part and the soft constraint part, improve the expression ability of the performance evaluation function, and also enhance the perception ability of the optimization process to the performance index conflict, so as to guide the search to jump out of the local optimum and tend to the optimal solution of the global performance of the system.

[0106] In the above embodiment, the performance difference between different sample controlled objects is not considered in the control parameter optimization process, which may lead to that the parameter combination obtained by optimization, i.e. the optimal candidate control parameter value combination, is only applicable to a specific sample controlled object, and cannot guarantee the universality and consistency of the parameters for different samples in batch production.

[0107] In an embodiment, the evaluation part in the performance evaluation function output by the above evaluation method based on each performance index and the index weight includes: outputting sample weights of each sample controlled object based on the gradient interval of the sample controlled object; and outputting the evaluation part in the performance evaluation function based on the evaluation method based on each performance index and the index weight, and the sample weights of each sample controlled object.

[0108] The gradient interval division can be based on the physical characteristic parameters of the sample controlled objects, such as the damping coefficient of the motor. Specifically, based on the gradient interval of the sample controlled object, the sample weight of each sample controlled object is output. For example, according to the damping coefficient of the motor, the damping coefficient is divided into three intervals of large gradient, medium gradient and small gradient, and each interval corresponds to different sample weights. The sample weight distribution follows the principle of giving priority to the medium gradient, for example, the sample weight of the medium gradient is set to 0.5, and the sample weights of the large gradient and the small gradient are set to 0.3 and 0.2 respectively. The construction of the evaluation part adopts a double weight superposition mechanism to linearly combine the performance index weight and the sample weight, for example, the evaluation value of a single sample controlled object is calculated as ∑(performance index score × index weight) × sample weight, and the final evaluation value is the sum of the weighted evaluation values of all sample controlled objects.

[0109] Specifically, in the motor parameter optimization of the optical stabilization system, first, the test samples are divided into three gradient intervals according to the damping coefficient, for example: the damping coefficient less than 0.05 N·s / m is classified into the small gradient interval, 0.05-0.15 N·s / m is the medium gradient interval, and greater than 0.15 N·s / m is the large gradient interval. Each interval of the sample is respectively given a weight coefficient of 0.2, 0.5 and 0.3. When determining the performance evaluation function, for each candidate control parameter value combination, the performance index score of the sample controlled object in each gradient interval is calculated, the performance index score is multiplied by the index weight, and then multiplied by the sample weight of the corresponding gradient interval. Finally, the average value of the weighted scores of all sample controlled objects is obtained. This double weight mechanism automatically adapts to the actual distribution characteristics of the samples in the parameter optimization process, focuses on optimizing the adaptability of the high proportion of medium gradient samples, and at the same time considers the performance of extreme samples, so as to obtain an optimal parameter combination with batch applicability.

[0110] Further, based on the evaluation method of each performance index, the index weight, and the sample weight of each sample controlled object, the evaluation part in the performance evaluation function is output. Specifically, a plurality of performance indicators can be set, such as overshoot, settling time, steady-state error, etc., and each indicator is assigned a weight. For example, the overshoot weight is 0.4, the settling time weight is 0.4, and the steady-state error weight is 0.2. Then, for each sample, the weighted performance score is calculated. Finally, the weighted performance scores of all samples are again weighted and averaged according to the sample weight to obtain the final evaluation result of the performance evaluation function, such as the evaluation score / performance evaluation value.

[0111] Correspondingly, in some embodiments, the evaluation part of the performance evaluation function can be specifically shown in the following formula (5).

[0112]

[0113] wherein, is the score of the i-th performance indicator of the j-th sample controlled object; w i is the corresponding indicator weight of the performance indicator, and the indicator weights of the performance indicators satisfy i i, M is the number of sample controlled objects, and N is the number of performance indicators. It should be noted that the sample weights of the sample controlled objects in formula (5) are the same.

[0114] In other embodiments, different sample weights of the sample controlled objects can be considered, and a performance evaluation function of formula (6) can be obtained. Wherein, k j represents the sample weight of the j-th sample controlled object.

[0115]

[0116] Embodiments of the present application consider the performance differences of different sample controlled objects, and improve the universality of control parameter optimization. By introducing a sample weight mechanism, the adaptability to medium gradient samples is focused on, while extreme samples are also considered, so that the finally optimized parameter combination can perform better stability and consistency in batch production. This double weight design not only guarantees the importance difference of each performance indicator, but also considers the distribution characteristics of different samples in actual production, effectively balancing the contradiction between single sample optimization and batch production demand.

[0117] Wherein, in some embodiments of the present application, the performance evaluation function of the plurality of performance indicators is as shown in formula (5) or formula (6), in these embodiments, the evaluation part and the sample controlled object are considered, but the constraint relationship between the performance indicators is not considered.

[0118] In other embodiments, formula (5) is only used as an evaluation part of the performance evaluation function, and the performance evaluation function also includes a constraint part. For example, in some embodiments, the performance evaluation function is as shown in formula (7).

[0119]

[0120] In other embodiments, formula (6) is only used as an evaluation part of the performance evaluation function, and the performance evaluation function also includes a constraint part. For example, in some embodiments, the performance evaluation function is as shown in formula (8).

[0121]

[0122] The formula not only considers the performance differences of different sample controlled objects, but also considers the constraint relationship between the performance indicators, so that the performance evaluation function is more accurate, and the finally determined control parameter value is also more accurate.

[0123] In an embodiment, asFigure 2 As shown in the above S104, i.e., the step of determining the control parameter value of the control member associated with the performance index based on the candidate control parameter value combination and the performance evaluation function, comprises the following steps:

[0124] S201, based on each candidate control parameter value combination and the performance evaluation function, output the performance evaluation value of each candidate control parameter value combination.

[0125] Each candidate control parameter value combination is verified by the verification item one by one to obtain the performance index score corresponding to each performance index, and then the performance index score is brought into the performance evaluation function to obtain the score result of the performance evaluation function, that is, the performance evaluation value of each candidate control parameter value combination. Each candidate control parameter value combination corresponds to a performance evaluation value.

[0126] S202, based on the performance evaluation value of each candidate control parameter value combination and the optimal performance evaluation value determination method, output the optimal performance evaluation value.

[0127] The determination method of the optimal performance evaluation value includes a preset determination logic, for example, including a highest total score strategy or a lowest score strategy. For example, the performance index score is brought into the performance evaluation function, and the highest performance evaluation value in the score result of the performance evaluation function, i.e., the performance evaluation value, is determined as the optimal performance evaluation value.

[0128] S203, the optimal control parameter value combination corresponding to the optimal performance evaluation value is determined as the control parameter value of the control member.

[0129] The number of optimal performance evaluation values may also appear multiple, when multiple optimal performance evaluation values appear, the candidate control parameter value combination with the highest performance index score of the core index performance index is preferred as the optimal control parameter value combination, thereby obtaining the control parameter value of the control member.

[0130] The embodiment of the present application improves the control parameter selection efficiency and accuracy, reduces manual trial and error, can be automatically used in laboratory or production line, further, the parameter universality is stronger, avoids single motor overfitting, and ensures the balance of each verification index, has strong adaptability, and can support different optical image stabilization systems.

[0131] In order to better implement the control parameter determination method of the embodiment of the present application, the embodiment of the present application also provides a control parameter determination device. Please refer to Figure 3 , Figure 3 The structure diagram of the control parameter determination device provided by the embodiment of the present application. The control parameter determination device 300 can include a first acquisition module 301, a first determination module 302, a second acquisition module 303, and a second determination module 304.

[0132] The first obtaining module 301 is configured to obtain at least one performance index of the optical image stabilization system, and determine a candidate control parameter associated with the performance index according to the performance index.

[0133] The first determining module 302 is configured to determine a candidate control parameter value combination according to the candidate control parameter.

[0134] The second obtaining module 303 is configured to obtain a performance evaluation function associated with the performance index of the optical image stabilization system.

[0135] The second determining module 304 is configured to determine a control parameter value of a control parameter associated with the performance index in the control member based on the candidate control parameter value combination and the performance evaluation function.

[0136] All the technical solutions described above can be combined to form optional embodiments of the present application, which will not be described one by one.

[0137] Correspondingly, the present application also provides an electronic device, which can be a terminal or a server. As shown in Figure 4 Figure 4 The electronic device provided in the embodiments of the present application is a structural schematic diagram. The electronic device 400 includes a processor 401 having one or more processing cores, a memory 402 having one or more computer readable storage media, and a computer program stored in the memory 402 and executable on the processor. Wherein, the processor 401 is electrically connected with the memory 402.

[0138] The processor 401 is the control center of the electronic device 400, which connects each part of the electronic device 400 by various interfaces and lines, executes various functions and processes data of the electronic device 400 by running or loading the software program (computer program) and / or module stored in the memory 402 and calling the data stored in the memory 402, so as to monitor the electronic device 400 as a whole.

[0139] In the embodiments of the present application, the processor 401 in the electronic device 400 will load the instructions corresponding to the process of one or more application programs / computer programs into the memory 402, and run the application program stored in the memory 402 by the processor 401, so as to realize various functions, such as the following functions.

[0140] ​At least one performance index of the optical image stabilization system is acquired, and a candidate control parameter associated with the performance index is determined according to the performance index; a candidate control parameter value combination is determined according to the candidate control parameter; a performance evaluation function associated with the performance index of the optical image stabilization system is acquired; and a control parameter value of the control member associated with the performance index is determined based on the candidate control parameter value combination and the performance evaluation function.

[0141] The specific implementation and achieved benefits of each operation above can be seen from the foregoing embodiments, which will not be described herein.

[0142] In an embodiment, the embodiments of the present application further provide a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the control parameter determination method described above. The non-transitory computer-readable storage medium has all the benefits of the control parameter determination method described above, which will not be described herein.

[0143] The computer-readable storage medium may, for example, be an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any suitable combination of the above, which is not specifically limited herein. More specific examples of the computer-readable storage medium can include, but are not limited to, an electrical connection having one or more wires, a portable computer diskette, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above.

[0144] In some embodiments of the present application, the computer-readable storage medium can be any tangible medium that contains or stores a program that can be used by or in conjunction with an instruction execution system, apparatus, or device. The computer-readable storage medium described above can be included in the electronic device described above, or can exist separately from the electronic device without being assembled into the electronic device.

[0145] In an embodiment, the embodiments of the present application further provide a computer program product comprising a computer program, which, when executed by a processor, implements the steps of the control parameter determination method described above. The computer program product has all the benefits of the control parameter determination method described above, which will not be described herein.

[0146] In the above embodiments, the description of each embodiment focuses on different aspects, and the parts not described in detail in a certain embodiment can be seen from the related description of other embodiments.

[0147] The embodiments, implementation manners, and related technical features of the present application can be combined or replaced with each other without conflict.

[0148] The above merely describes preferred embodiments of the present application, and is not intended to limit the present application in any form. Although the description of each embodiment in the present application has its own emphasis, the parts not described in detail in a certain embodiment can be referred to the relevant description of other embodiments. Any simple modification, equivalent change and modification made to the above embodiments according to the technical essence of the present application without departing from the technical solution content of the present application still falls within the scope of the technical solution of the present application.

Claims

1. A method for determining control parameters, characterized in that, For determining at least one control parameter value of a control component in an optical image stabilization system, including: Obtain at least one performance index required by the optical image stabilization system, and determine the associated candidate control parameters based on the performance index; Determine the combination of candidate control parameter values ​​based on the candidate control parameters; Obtain the performance evaluation function associated with the performance index of the optical image stabilization system; Based on the candidate control parameter value combination and the performance evaluation function, the control parameter values ​​associated with the performance index in the control component are determined.

2. The method for determining control parameters according to claim 1, characterized in that, The step of determining the candidate control parameter value combination based on the candidate control parameters includes: Based on the range of candidate parameter values ​​for each candidate control parameter, output parameter space; Based on the parameter selection method, the candidate control parameter value combination is output according to the parameter space.

3. The method for determining control parameters according to claim 2, characterized in that, The parameter selection method includes a search method, and the step of outputting the candidate control parameter value combination according to the parameter space based on the parameter selection method includes: Based on the search strategy corresponding to the search method, the candidate control parameter value combination is output according to the parameter space; or, The parameter selection method includes an optimization method. The step of outputting the candidate control parameter value combination based on the parameter selection method and the parameter space includes: Based on the initial parameter values ​​and optimization strategy corresponding to the optimization method, the candidate control parameter value combination is output according to the parameter space.

4. The method for determining control parameters according to claim 1, characterized in that, The method for obtaining the performance evaluation function associated with the performance index of the optical image stabilization system includes: Based on the evaluation method and weight of each performance indicator, the evaluation part of the performance evaluation function is output. Based on the constraint relationships of each performance index, the constraint part of the performance evaluation function is output; The performance evaluation function associated with the performance index of the optical image stabilization system is determined based on the evaluation unit and the constraint unit.

5. The method for determining control parameters according to claim 4, characterized in that, The evaluation methods and weights of each performance indicator are determined as follows: Based on the optical image stabilization attention index, output the performance index and the index weight of each performance index used to form the performance evaluation function; Based on the index values ​​and quantification methods of each performance index, the evaluation method for each performance index is output.

6. The method for determining control parameters according to claim 4, characterized in that, The evaluation part of the performance evaluation function, based on the evaluation method and weights of each performance indicator, includes: Based on the gradient range of the controlled sample, output the sample weight of each controlled sample; Based on the evaluation method and weight of each performance index, and the sample weight of each controlled object, the evaluation part of the performance evaluation function is output.

7. The method for determining control parameters according to any one of claims 1 to 6, characterized in that, The determination of the control parameter values ​​of the control device based on the combination of candidate control parameter values ​​and the performance evaluation function includes: Based on the candidate control parameter value combinations and performance evaluation function, output the performance evaluation value of each candidate control parameter value combination; Based on the performance evaluation value of each candidate control parameter value combination and the method for determining the optimal performance evaluation value, the optimal performance evaluation value is output. The optimal control parameter value combination corresponding to the optimal performance evaluation value is determined as the control parameter value of the control component.

8. The method for determining control parameters according to claim 1, characterized in that, The step of determining the associated candidate control parameters based on the performance indicators includes: Obtain the type of the control component; determine a first candidate control parameter associated with the performance index from the control parameters that match the type of the control component; based on the first candidate control parameter, determine a second candidate control parameter associated with the performance index from the control parameters that match the type of the control component; use the first candidate control parameter and the second candidate control parameter as candidate control parameters; or, The type of the control element is obtained, the control element includes a first control element and a second control element; a third candidate control parameter associated with the performance index is determined from the control parameters that match the type of the first control element; based on the third candidate control parameter, a fourth candidate control parameter associated with the performance index is determined from the control parameters that match the type of the second control element; the third candidate control parameter and the fourth candidate control parameter are used as candidate control parameters.

9. An electronic device, characterized in that, include: A memory on which computer programs are stored; A processor for executing the computer program in the memory to implement the control parameter determination method according to any one of claims 1 to 8.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the control parameter determination method according to any one of claims 1 to 8.

Images