52 results about "Algorithm robustness" patented technology

The invention discloses an outdoor robot path planning method based on improved bidirectional A * and DWA algorithms. The method comprises the following steps: step 1, A * algorithm improvement; step 2, DWA algorithm improvement; step 3, algorithm fusion; step 4, path planning; the dynamic weight DOMW strategy is introduced into the A * algorithm to optimize a heuristic function, multiple factors are comprehensively considered, path planning is more reasonable, the algorithm calculation amount and search time are reduced by removing redundant nodes, the algorithm efficiency is improved, a bidirectional search mechanism is adopted, and the algorithm efficiency is improved. The search efficiency is improved, the search space is reduced, a local optimal solution is avoided, the robustness of the algorithm is enhanced, the path is smoothed by adopting a Bezier curve optimization strategy, so that the path meets the actual motion requirement of the robot, the damage to hardware is reduced, and the attractiveness and readability of the path are improved; by optimizing a DWA algorithm trajectory evaluation function and speed space constraints, the collision risk is reduced, and local obstacle avoidance and operation stability are improved.

The invention discloses a rapid local repair method for a self-intersecting triangular mesh. According to the method, local processing and rational number calculation are combined, aiming at an input low-quality triangular mesh, firstly, the mesh quality is optimized through a preprocessing step, intersection areas are positioned, then the mesh is partitioned based on octree division and connected component analysis, and the intersection areas and non-intersection areas are separated. In each intersection area, a shape-preserving boundary three-dimensional grid optimization method based on rational numbers is adopted to ensure the robustness of intersection calculation, and surface grids are extracted after successful conversion into floating-point number representation. And finally, seamlessly splicing the repaired local grid and the non-intersecting grid, and improving the overall grid quality through an edge re-gridding method based on curvature self-adaption. According to the method, the non-intersecting high-quality grid can be quickly generated, the method is suitable for downstream applications such as finite element analysis, meanwhile, the calculation overhead is remarkably reduced, and the algorithm robustness is improved.

The invention relates to the field of power optical cable facilities, in particular to a power optical cable fault detecting and positioning method based on an extreme learning machine. The OTDR signal is subjected to multi-scale decomposition and de-noising processing based on wavelet packet analysis, and high-frequency noise is effectively suppressed and key event point features are reserved by refining frequency band division, so that the signal quality is improved. And constructing a fault classification model based on an extreme learning machine, and realizing efficient classification and fault mode identification of de-noised OTDR data. And in combination with a white whale optimization algorithm (BWO), carrying out optimization search on a fault detection result output by the classification model so as to improve the detection precision and algorithm robustness in a complex optical cable environment. And an improved Dijkstra algorithm is introduced to realize accurate positioning and path tracking, so that the accuracy and response speed of fault positioning are remarkably improved. Through multi-level algorithm fusion, the problems of signal noise interference, insufficient pattern recognition accuracy and low positioning precision in a complex network environment in optical cable fault detection are solved.

The invention discloses a cognitive interference decision-making method based on deep reinforcement learning and an application system of the cognitive interference decision-making method, which are suitable for an interference unmanned aerial vehicle to complete an intelligent interference decision-making task when resisting a networking radar system in a complex and changeable electromagnetic environment. In order to improve the performance of an existing interference decision-making system in the aspects of sample efficiency and strategy stability, the interference decision-making algorithm DAR-PPO comprises two major innovation modules: one module is a KTD-Clip module which constructs a feedback signal based on experience KL divergence and dynamically adjusts a clipping threshold value in strategy updating so as to control the strategy updating step length, so that the strategy updating step length is always in a stable and effective trustworthiness region; the adaptive capacity of the algorithm to environment change is obviously improved; the second module is an RA-IS module, effective reuse of historical interaction samples is achieved by constructing a trajectory playback pool and introducing clipping importance sampling, meanwhile, a weight entropy regularization mechanism is designed, strategy fluctuation caused by abnormal weights is restrained, and therefore the training sample utilization rate is remarkably increased, and the algorithm robustness is enhanced.

The invention relates to the technical field of cloud computing and distributed computing scheduling, in particular to a server-free MapReduce job scheduling optimization method. Comprising the following steps: initializing Bayesian genetic algorithm operation parameters; forming an initial population; the current population executes variable neighborhood search to generate a new solution, and the population is updated; constructing a Bayesian probability model; generating a new solution through Bayesian probability sampling and genetic manipulation, and updating the population; updating a global optimal solution; and judging whether the time limit is reached or not, if so, outputting a globally optimal solution and the corresponding maximum completion time, and if not, continuing iteration. The application of the method has the positive effects of minimizing the maximum completion time of the operation and improving the robustness of the algorithm and the scheduling efficiency.

The invention discloses a competitive evolution multi-task optimization method, system and equipment, and aims to solve the problem that an existing method is difficult to carry out collaborative optimization on fuel cost and gas emission under the condition that power balance and unit capacity constraint conditions are met. The method comprises the following steps: constructing main and auxiliary double-population parallel search; state indexes representing population convergence, diversity and feasibility are calculated in real time; based on the current state, a cooperation mode and evolution operator combined action is adaptively selected through a deep reinforcement learning agent; rewards are calculated according to improvement of the population on cost, emission and constraint satisfaction after action execution; and training the intelligent agent by using the state, the action and the reward, and iteratively optimizing the decision strategy until the Pareto optimal scheduling scheme meeting the constraint is output. According to the method, adaptive intelligent guidance of the evolutionary process is realized, and the optimization quality, the convergence speed and the algorithm robustness of the economic emission scheduling scheme of the power system are remarkably improved.

The invention relates to the technical field of power electronics and direct current transmission, in particular to a two-level VSC-HVDC minimum direct current capacitance calculation method based on LMI. The method comprises the following steps: establishing a state equation comprising a main circuit and a controller dynamic model, converting a three-phase alternating current quantity into a direct current quantity under a synchronous rotating coordinate system by utilizing Park transformation, and deducing a system state equation in combination with a Kirchhoff's law; performing Taylor expansion and linearization on the nonlinear equation, and constructing a small signal model based on a Lyapunov stability theory; by separating a diagonal matrix and defining a stability criterion in an LMI form, minimum direct-current capacitance solving is converted into a convex optimization problem, a critical value is iteratively searched by adopting a dichotomy method to determine minimum capacitance, and meanwhile, a tolerance mechanism is introduced to improve the robustness of an algorithm. Through combination of mathematical modeling and an optimization algorithm, the calculation efficiency and accuracy are effectively balanced, and a theoretical basis is provided for reducing the cost and the size of the converter.

The invention discloses a virtual assembly method and system based on improved point cloud registration and precision feature extraction, relates to the technical field of virtual assembly, and aims to solve the technical problem that a conventional ICP algorithm is insufficient in robustness and precision in a point cloud registration link in a current virtual assembly technology based on point cloud data. Comprising a file processing module, a preprocessing module, a point cloud registration module and a many-to-many component matching scheme solving module. According to the method, the RICP algorithm is designed, and a general adaptive robust function is introduced, so that the problem that the traditional ICP algorithm is sensitive to noise and outliers is effectively solved. The robust function can dynamically give a weight according to a point pair distance, stable and high-precision registration can still be realized even in a scene of large point cloud initial pose difference, low overlapping degree or unobvious surface features, and the problem of insufficient robustness and precision of a traditional ICP algorithm in a point cloud registration link of a current virtual assembly technology based on point cloud data is solved.

The invention discloses a three-stage AC optimal transmission switching method based on a relaxation model, and the method comprises the following steps: 1, screening a breakable branch through solving an optimal solution of the relaxation model in a first stage, and reducing the scale of a system; 2, in the second stage, sorting is carried out by solving the optimal solution of the alternating-current optimal power flow of disconnection of a single branch; and step 3, in the third stage, solving by adopting a branch and bound method according to the sorting sequence of the second stage, and finally determining an optimal solution of alternating current optimal transmission switching and a branch on-off scheme. According to the method, the dimensionality of the optimal transmission switching problem is effectively reduced, and the calculation efficiency and speed are improved; an alternating current optimal transmission switching relaxation problem is solved by adopting a method based on a quotient gradient system and IPOPT, so that the calculation convergence and speed are guaranteed, and the algorithm robustness is improved; branch and bound calculation is carried out according to the alternating current optimal power flow solution sorting result sequence of the cut-off single branch, and it is guaranteed that a high-quality line cut-off scheme is obtained through solving; in the calculation process, island branch processing is considered, the island phenomenon in the topology change process is prevented, and power grid splitting and unnecessary load outage are effectively avoided.

The invention discloses a beam focusing method and system based on a liquid neural network and meta learning, and belongs to the technical field of wireless communication, and the method comprises the following steps: determining a system model; determining an optimization target; and joint optimization. The LNNM-BF algorithm with high robustness and unsupervised learning is provided, the LNNM-BF algorithm designs a phase shift matrix theta through an index matrix S and a phase compensation matrix phi, then an analog precoding matrix FPS and a digital precoding matrix FBB are designed at a BS end, and the near-field dual-beam splitting effect can be well relieved without introducing an extra true time delayer. Besides, the LNNM-BF algorithm enhances the adaptability of beam focusing to imperfect CSI through a meta-learning framework and improves the robustness of the algorithm, and the embedded LNN dynamically adjusts the topological structure and the connection weight between neurons through the currently input and previously implied iteration information to realize the maximization of SE.

This invention relates to a 3D UAV path planning method based on an improved gray wolf optimization algorithm, comprising: acquiring a 3D environment model; setting the start and end positions of the UAV based on the 3D environment model; employing the improved gray wolf optimization algorithm, using the objective function as the fitness function; updating the positions of all gray wolves synchronously during the iterative process; and outputting the optimal path. This invention introduces a circle chaotic mapping to the gray wolf optimization algorithm to increase population diversity; introduces an adaptive exploration factor to increase the balance between global and local search; combines the global and local development formulas of the dung beetle optimization algorithm to compensate for the slow convergence speed of the traditional gray wolf optimization algorithm; subsequently introduces a sine and cosine strategy to help the algorithm search for potential optimal solution regions, thus better exploring the global optimal solution; and finally introduces a pooling mechanism to randomly recombine stored historical poor solutions to generate new solutions, accelerating the convergence speed and improving the robustness of the algorithm.

The present invention provides a robustness evaluation method and system for a data stream local differential privacy algorithm. The method judges the robustness of the algorithm by the proximity between the algorithm release frequency distribution at each timestamp and the algorithm robustness evaluation target distribution. The method adopts a branch judgment mechanism, which can adaptively judge at each timestamp which strategy the data stream local differential privacy algorithm enters to achieve a better evaluation effect for the robustness evaluation of the algorithm. Based on the judgment of the branch judgment mechanism, the method affects the dissimilarity calculation module of the data stream local differential privacy algorithm so that the privacy algorithm can enter the strategy expected by the evaluator, and is divided into two types: input evaluation and output evaluation. If the data stream local differential privacy algorithm enters the release strategy, the method provides a fine-grained algorithm robustness evaluation method for the frequency statistics local differential privacy algorithm in the release strategy. The method is also divided into two types: input evaluation and output evaluation.

The invention relates to the technical field of low-earth-orbit satellite communication, and provides a low-earth-orbit satellite edge calculation unloading method based on an attention mechanism, which comprises the following steps of: 1, generating a 500000-group sample set; 2, establishing a low earth orbit satellite moving edge calculation unloading model, and constructing a low earth orbit satellite calculation unloading objective function and constraint; 3, defining a Markov decision process tetrad in reinforcement learning; 4, constructing an unloading strategy solving model based on the attention network; and 5, setting a gradient descent method training parameter and a loss function, and optimizing a network parameter. According to the method, the edge calculation unloading strategy is formulated for the low-orbit satellite through the attention network, the problem that algorithm convergence needs to be achieved by greatly increasing the number of algorithm iterations during large-scale user unloading decision making is solved, and the method has the advantages of being high in algorithm robustness and task planning speed; the edge unloading decision making time of large-scale users is shortened, and the resource utilization rate of the system is optimized.

The invention provides a hybrid optimization algorithm-based steel plate stack transfer sorting system and method and a storage medium. The method comprises the steps of firstly, obtaining a to-be-sorted steel plate set, the number of available temporary stack positions and a target sequence; generating an initial stack transfer strategy through a heuristic rule; secondly, performing iterative optimization on the initial stack transfer strategy by using a genetic algorithm to generate an optimal stack transfer strategy by taking the minimum operation time length and the number of temporary stack positions used for stack transfer sorting as targets; and finally, stack transferring equipment is controlled to execute the stack transferring step in the optimal stack transferring strategy, and moving operation of the steel plates between the stack positions is achieved till stack transferring sequencing is completed. According to the method, the heuristic rule, the GA and the DQN are fused, so that the ternary relationship among the time constraint, the equipment resource and the temporary stack position can be coordinated, and the idling rate of the equipment is reduced, so that the production process of a shipyard is optimized, the production efficiency is improved, and meanwhile, the production cost is reduced; according to the method, the limitation of a traditional single method is broken through, the algorithm robustness is higher, and the industrial application value is prominent.

The present invention discloses a point set registration method based on continuously differentiable least trimmed squares estimation. The present invention proposes a continuously differentiable least trimmed squares criterion based on the least squares criterion and the least trimmed squares criterion. The cost function of this criterion is continuous and differentiable, so it can be more easily applied to gradient-based learning strategies, solving the defect that the least trimmed squares criterion is not differentiable. The present invention applies the continuously differentiable least trimmed squares criterion to the point set registration problem, which can better handle the point set registration situation in the presence of outliers, making the registration process as little affected by outliers as possible, so as to obtain a relatively accurate registration result. By using the continuously differentiable least trimmed squares criterion, the present invention can effectively avoid the problem that the robustness of the algorithm is sensitive due to improper selection of the trimming parameter. The present invention avoids the need for an additional sorting step in the traditional LTS criterion-based registration method, thus simplifying the algorithm flow and making the application more convenient.

The present invention discloses a method for stable control of an unmanned aerial vehicle against wind disturbances. Based on the traditional active disturbance rejection control, it is optimized by fusing the Kalman filter KF and the extended state observer ESO to form a composite state observer KFESO. The ESO can ensure the operation of the KF when there are deviations in the system or uncertainties in the model, and the KF can filter the input of the ESO to weaken the influence of noise on the ESO. At the same time, the backstepping method is added to the active disturbance rejection controller, which can improve the response speed of the system, enabling the unmanned aerial vehicle system to respond more quickly when affected by wind disturbances and enhancing the stability of the system. The improved active disturbance rejection controller is adopted in both the position loop and the attitude loop in order to achieve better anti-interference performance and ensure a high response speed of the unmanned aerial vehicle under wind disturbances. The present invention improves the robustness of the algorithm, simplifies the algorithm design, improves the calculation efficiency, enhances the adaptability and generalization ability, and promotes the practical application ability.

The invention discloses an LMI-based intelligent suspension robust control and state observation method, and relates to the technical field of suspension control, and the method comprises the following steps: S1, splitting a control system into a non-matching disturbance subsystem and a matching disturbance subsystem; s2, determining state feedback control of an LMID algorithm according to the weight matrix of state control; determining disturbance compensation of a disturbance observer part in an LMID algorithm based on a weight matrix of a matching disturbance subsystem and disturbance estimation, and realizing the LMID algorithm by using an LMIob observer; and S3, taking the sprung acceleration based on the IMU in the suspension system as the input quantity of the LMIob observer, and obtaining state estimation of the control system. The method is designed based on an LMID algorithm of LMI, the algorithm has disturbance observation properties, the robustness of the algorithm can be improved, a weight matrix is introduced, and flexible adjustment of state control and disturbance estimation is achieved.

The invention discloses a mining subsidence prediction parameter solving method based on an improved center collision optimization algorithm, and belongs to the field of mine deformation monitoring data processing. Aiming at the defects that a basic center collision optimization algorithm is prone to premature convergence and insufficient in optimization precision when solving a mining subsidence parameter inversion problem, a Huber loss function is adopted to construct a fitness evaluation model to enhance the robust capability of the algorithm, and Cubic mapping is utilized to generate a random number to improve the stability of a search process, so that the method is suitable for the mining subsidence parameter inversion problem. And a self-adaptive elite-guided Cauchy variation mechanism is introduced to enhance the global exploration capability. According to the method, firstly, a probability integral method parameter inversion problem is constructed into an optimization model, and a population is initialized; performing double-space collaborative search in an original space and a decorrelation space constructed based on principal component analysis, and iteratively updating a population in combination with a dynamic space allocation strategy; and finally outputting an optimal parameter solution. According to the method, the convergence precision, stability and robustness of the algorithm in complex nonlinear parameter inversion are effectively improved.

The present application relates to the technical field of automatic driving model test verification, and is a hardware-in-the-loop automatic driving model vulnerability mining method and system, specifically comprising: in the hardware-in-the-loop closed loop operation, through the non-invasive probe deployed in the tested chip, the neuron activation features of the model hidden layer are extracted, and the cognitive uncertainty index is calculated; when the uncertainty exceeds the threshold value, the physical parameters such as illumination, weather, shielding and target geometric shape are dynamically adjusted, and the targeted adversarial negative sample scene is generated; in the scene evolution process, the algorithm output, the model internal state and the chip hardware state are monitored synchronously, the failure reason is determined through multidimensional comparison, and the vulnerability is classified into algorithm logic weakness, quantization precision weakness or hardware performance weakness. The present application carries out active exploration based on cognitive driving, realizes efficient and accurate mining of model vulnerability and finding of software and hardware coupling failure, and is especially suitable for algorithm robustness evaluation of domestic heterogeneous computing platforms.

The present invention discloses a short-term prediction method for railway traction load, which belongs to the field of power data processing technology. The present invention includes data cleaning, data primary decomposition, data secondary decomposition, sample entropy clustering, model training, and data prediction. The present invention combines the intrinsic time scale decomposition algorithm, the robust local mean decomposition algorithm, sample entropy clustering, the time convolution network, and the bidirectional gated recurrent network to establish a short-term prediction method for traction load based on a combination of secondary decomposition reconstruction and deep learning double-layer optimization. This method can reduce the complexity of high-frequency components. At the same time, it effectively solves the defect that existing research cannot deeply capture the real-time dynamic characteristics of the load, thereby improving the accuracy of traction load prediction.

The application provides an improved tabu search topology optimization method of a marine wind farm power collection system, which is based on the positions of wind turbines to intelligently preliminarily group and sort, to form an initial connection scheme, to connect all the wind turbines in each group with the minimum cost, to calculate the group cable cost, to check whether the lines cross, to impose a penalty in the total cost if the lines cross, to ensure that the scheme is feasible, to compare the current scheme with the historical best, to update the record and intelligently exchange the wind turbine groups to explore a new scheme, to repeat the optimization process until the number of iterations is satisfied, and to finally output a globally optimal topology scheme. The application has the advantages of better optimization effect, significantly improved economy, higher calculation efficiency, stronger engineering practicability, better algorithm robustness, better scheme reliability, high design automation degree and strong auxiliary decision-making capacity.

The application relates to a non-cooperative spacecraft active tracking method based on deep reinforcement learning, belongs to the field of aerospace, and aims to solve the defects that existing algorithms cannot fuse a spacecraft dynamics model and satellite orbit dynamics, cannot effectively improve algorithm robustness on the basis of ensuring tracking precision, and cannot effectively extract time sequence related information about a target from training samples. The process is as follows: 1, obtaining the position and speed of a pursuing spacecraft; 2, constructing an actor network and a critic network and a loss function; 3, constructing a reward function, obtaining a trained actor network and critic network; 4, obtaining the difference between the position and the expected position between a target spacecraft and the pursuing spacecraft, and the speed difference between the target spacecraft and the pursuing spacecraft, inputting the trained actor network, and outputting the action of the current time step by the trained actor network.

The invention relates to the technical field of ship control and simulation, in particular to a physical isolation-based ship control algorithm robustness verification method, which comprises the following steps of: constructing a ship model at an embedded end, and combining and establishing the ship model based on an MMG ship operation model for determining an inertia matrix parameter and a linear damping coefficient; the embedded terminal is connected with an upper computer, the upper computer issues a motion instruction to the ship model, the ship model outputs ship motion state data based on the motion instruction, ship type parameters and a sea condition disturbance scene, and the embedded terminal transmits the ship motion state data to the upper computer; and the upper computer compares the ship motion state data with a preset track, and evaluates the robustness of the ship control algorithm under different disturbances. According to the method, the execution effect of the control strategy on resource-limited hardware can be accurately reflected, and a reliable experimental foundation and engineering support are provided for testing the robustness of ship path tracking and attitude control research under complex sea conditions.

The present application relates to the technical field of cloud computing and distributed computing scheduling, and particularly belongs to a kind of serverless MapReduce job scheduling optimization method, comprising: initializing the operating parameter of bayesian genetic algorithm;Form initial population;Current population executes variable neighborhood search to generate new solution, update population;Bayesian probability model is constructed;New solution is generated by bayesian probability sampling and genetic operation, and population is updated;Global optimal solution is updated;Whether to reach time limit is judged, if reach, then the global optimal solution and its corresponding maximum completion time are output, if not reach, then continue iteration.The application has the positive effect of realizing the minimum maximum completion time of job and improving the robustness of algorithm and scheduling efficiency.

The invention relates to an intelligent algorithm application range analysis method based on region growth, and belongs to the technical field of intelligent algorithm robustness evaluation. According to the method, the capability boundary of an intelligent algorithm is iteratively generated by expanding a region growing and automatic searching method, firstly, a seed set is formed based on a targeted evaluation sample for preliminary evaluation, secondly, the capability boundary is iteratively described continuously by adopting a region growing method based on sample disturbance, and finally, contour lines are formed based on regression methods such as Kernel and the like. The safe distance is obtained through calculation, and support and guidance are provided for optimization iteration and landing application of an intelligent algorithm.

The invention discloses a well track automatic obstacle avoiding method for guiding type fence searching. The method is characterized by comprising the following steps that S1, a searching fence environment is built; s2, initializing an algorithm; and S3, node searching: S4, smoothing a search path to obtain an optimal and smooth path. By building a multi-layer fence search environment, parallel operation of an algorithm (multiple layers search paths at the same time) is facilitated, so that the overall search process is accelerated; the provided guiding type heuristic function is beneficial to guiding search to be carried out to areas not occupied by obstacles, the search space is reduced, the decision-making time in the search process is shorter, and the algorithm robustness is higher.

The application discloses a kind of competitive evolution multitask optimization method, system and equipment, to solve the problem that existing method is difficult to carry out collaborative optimization to fuel cost and gas emission under meeting power balance and unit capacity constraint condition.The method comprises the following steps: constructing main, auxiliary double population parallel search;Real-time calculation state index representing population convergence, diversity and feasibility;Based on the current state, through deep reinforcement learning agent self-adapting selection cooperation mode and evolution operator combination action;According to the improvement of population after action execution in cost, emission and constraint satisfaction, calculate reward;State, action and reward are used to train agent, iteratively optimize decision strategy, until the pareto optimal scheduling scheme that meets the constraint is output.The application realizes the adaptive intelligent guidance of evolution process, significantly improves the optimization quality, convergence speed and algorithm robustness of power system economic emission scheduling scheme.

The present application relates to the technical field of micro-grid optimization, and discloses a micro-grid optimization scheduling method and system based on an improved electric eel foraging optimization algorithm, which comprises constructing a target function with the minimum total operation cost of a micro-grid as the target, and establishing constraint conditions that the micro-grid needs to satisfy during operation; the micro-grid optimization scheduling model composed of the target function and the constraint conditions is solved based on the improved electric eel foraging optimization algorithm to generate an optimal scheduling scheme; the algorithm of the present application restructures the optimization mechanism in depth: in the initialization stage, an elite reverse learning strategy is adopted to mine potential high-quality solutions, significantly enhancing population diversity; in the migration and hunting stage of core evolution, a Cauchy mutation operator is introduced in combination with a tournament selection mechanism to improve the robustness of the algorithm while effectively avoiding premature convergence; in addition, a vertical and horizontal crossover strategy is fused to dynamically coordinate the algorithm behavior, achieving a precise balance between local search precision and global development capability.

The invention discloses a beam focusing method and system based on a liquid neural network and meta learning, and belongs to the technical field of wireless communication, and the method comprises the following steps: determining a system model; determining an optimization target; and joint optimization. According to the invention, the LNNM-BF algorithm with strong robustness and unsupervised learning is provided, the LNNM-BF algorithm designs a phase shift matrix through an index matrix and a phase compensation matrix, then an analog precoding matrix and a digital precoding matrix are designed at a BS end, and the problem of a near-field dual-beam splitting effect can be well relieved without introducing an extra true time delayer. Besides, the LNNM-BF algorithm enhances the adaptability of beam focusing to imperfect CSI through a meta-learning framework and improves the robustness of the algorithm, and the embedded LNN dynamically adjusts the topological structure and the connection weight between neurons through the currently input and previously implied iteration information to realize the maximization of SE.

The invention discloses a wavelet noise reduction method based on two-parameter optimization, relates to the technical field of digital signal processing, and solves the technical problems that in a traditional wavelet noise reduction method, selection of a threshold value and the number of decomposition layers depends on experience, and the noise reduction effect is unstable. According to the technical scheme, the method is characterized by comprising the steps of collecting noisy signals, performing multilayer wavelet decomposition, constructing a signal-to-noise ratio and root-mean-square error two-parameter evaluation index, performing joint optimization on a threshold value and the number of decomposition layers by adopting a self-adaptive optimization algorithm, performing wavelet threshold value noise reduction according to optimized parameters, and reconstructing the signals. According to the method, the optimal noise reduction parameter can be adaptively determined, the signal noise reduction quality and the algorithm robustness are remarkably improved, and the method is suitable for signal preprocessing in the fields of power system monitoring, mechanical vibration analysis, voice processing and the like.