12results about How to "Improve optimization accuracy" patented technology

The invention discloses an energy management balance scheduling optimization method and system based on deep learning, and the method comprises the steps: collecting multi-source feature data, carrying out the standardization and preprocessing, and obtaining structured multi-dimensional time series data; constructing a node relation graph by using an improved dynamic graph attention model, and generating a global attention matrix; node features are updated according to the global attention matrix, optimized features are formed, and a scheduling strategy is generated; optimizing the scheduling strategy by adopting an improved Harlisia eagle algorithm to obtain an optimized scheduling scheme; a node operation strategy is adjusted according to the optimization scheme, and power and load distribution is completed; and a final operation scheme is generated and dynamically corrected. According to the method, the improved dynamic graph attention model is constructed, and the Harris eagle optimization algorithm combining a boundary escape mechanism and a dynamic energy regulation and control mechanism is combined, so that intelligent modeling of multi-source data of an energy network, multi-target optimization of a scheduling strategy and dynamic balance scheduling of energy operation are realized.

This invention discloses an optimization design method for S-curve nozzles used in aero-engines that considers structural deformation, belonging to the field of aero-engines. The method selects key geometric parameters of the S-curve nozzle as characteristic parameters, chooses the structural deformation value at a characteristic location and two aerodynamic performance parameters as objective functions, then conducts orthogonal experimental design and constructs a response surface model based on the characteristic parameters and objective functions. After each optimization step, a cyclical evaluation is performed to obtain the final optimization result. This invention solves the problem of high aerodynamic performance loss caused by large structural deformation in S-curve nozzles used in aero-engines in existing technologies.

This invention discloses a software code optimization method and system based on big data. The method includes the following steps: constructing a code optimization knowledge base; acquiring the target code to be optimized and its corresponding estimated runtime data features; parsing the target code, extracting its structural features, and fusing the structural features with the estimated runtime data features to generate a target code fingerprint; performing similarity matching between the target code fingerprint and historical code fingerprints in the code optimization knowledge base to determine target historical code fingerprints that match the target code fingerprint; and determining a target optimization strategy for the target code based on the optimization strategy corresponding to the target historical code fingerprint. This invention can improve development efficiency, reducing the performance tuning work of senior architects, which can take hours or even days, to automated recommendations at the minute level, greatly improving software iteration efficiency.

This invention relates to a hierarchical optimization method for positive and negative pulse charging of lithium batteries based on deep reinforcement learning (DRL) and model predictive control (MPC), belonging to the field of battery management. This method establishes a long-term decision layer based on deep reinforcement learning and a short-term optimization layer based on model predictive control. The long-term decision layer learns the mapping relationship between positive and negative pulse charging actions and battery states across the entire process of positive and negative pulse charging, forming a long-term feasible region describing the overall charging trend, which is then passed to the short-term optimization layer in the form of a policy sequence. Near the long-term feasible region, the short-term optimization layer optimizes the positive and negative pulse charging actions on a short-term scale based on battery safety constraints and the current state, achieving optimal real-time control. This invention combines the efficient search capability of DRL in the nonlinear policy space with the rolling optimization mechanism of MPC under electrochemical constraints, realizing optimized control of the positive and negative pulse charging process of lithium batteries.

This invention discloses a method and system for coordinated scheduling of two-way arbitrage and demand response by energy storage aggregators. Through an innovative business model, it enables energy storage aggregators to achieve two-way arbitrage in both wholesale and retail markets, as well as multiple value acquisitions through collaborative demand response with industrial and commercial parks. This invention constructs a mixed-integer linear programming optimization model with the objective of maximizing the aggregator's total day-ahead revenue, accurately characterizing the aggregator's trading behavior and energy storage operation constraints in a multi-market environment. Addressing the nonlinear effective response capacity segmentation calculation and evaluation mechanism in demand response rules, this invention proposes a piecewise linearization method based on auxiliary binary variables and the Big-M method, accurately transforming complex nonlinear constraints into standard linear constraints. This method can effectively coordinate the optimal allocation of energy storage among multiple revenue sources, including wholesale market price arbitrage, retail market revenue sharing, and demand response compensation, significantly improving the economic benefits of energy storage aggregators and providing a scientific decision support tool for the commercial operation of energy storage aggregators.

The invention relates to the technical field of recycled asphalt mixture design, and discloses a recycled asphalt mixture design method and device, computer equipment and a storage medium, and the method comprises the following steps: constructing a grading modulus data set; determining a grading optimization domain; constructing a collaborative optimization mechanism; the design model is based on a target modulus, a collaborative optimization mechanism is combined, individuals of an adaptive target are iteratively screened in a grading optimization domain according to a gradient strategy of first modulus primary screening and then cost-working condition double-constraint fine screening, and an optimal grading result with the target modulus reaching the standard, the cost being controllable and the working condition being adaptive is output. The device, the computer equipment and the storage medium all correspond to the method. According to the method, the optimization precision of the high-aging RAP and the grading fluctuation aggregate is improved through the collaborative optimization mechanism of the grading modulus data set and the adaptive reclaimed material characteristics, and the gradient screening strategy is combined, so that it is guaranteed that the target modulus reaches the standard, cost controllability and working condition adaptation are considered, the trial preparation cost is reduced, and the reclaimed material utilization rate and the engineering practicability of the design scheme are improved.

The embodiment of the application provides a parameter optimization method for an outer contour surface of a door glass and an electronic device, the method comprising: obtaining initial control point position information, reference point position information and a target gap range, wherein the reference point position information is used to represent a space reference position corresponding to the outer contour surface of the door glass, and the target gap range is used to represent a gap constraint condition corresponding to the reference point position information; constructing an initial outer contour surface corresponding to the door glass based on the initial control point position information; performing evaluation processing on the initial outer contour surface by using the reference point position information and the target gap range to generate a modeling and engineering evaluation score; and performing iterative optimization on the initial control point position information based on the modeling and engineering evaluation score to obtain target control point position information, wherein the target control point position information is used to construct a target outer contour surface corresponding to the door glass. The application solves the technical problems of long development cycle and low optimization precision of related door glass large surface optimization methods.

The present application mainly relates to the technical field of automobile engineering. The present application discloses a kind of automobile wire harness full life cycle cost optimization method, the method includes: the multi-source data of integration component selection, design, assembly and logistics stage is generated cost base data table, breaks the data island of each link;Identify the optimizable item of each cost optimization dimension, generate cost reduction proposal scheme;Filter target cost reduction proposal scheme and combine, calculate the cost saving amount of each combination scheme, select the optimal cost reduction scheme;The optimal scheme is output to the execution end, and receive execution result data for iterative optimization.This scheme realizes the full life cycle cost optimization through the combination optimization and closed-loop feedback of global perspective, ensures the synergy of each link cost measure, avoids the overall cost rise caused by local optimization, and finally realizes the full life cycle cost optimization.

The invention discloses a graphic rendering performance optimization method and system for a high-dimensional configuration space, and belongs to the technical field of graphic rendering. According to the method, firstly, configuration parameters of a rendering system are optimized and modeled into a multi-objective optimization problem, and an initial population is generated; evaluating a performance index value of each individual by using a performance prediction model, and forming a target function vector in combination with a penalty function; a tournament selection strategy based on crowding distance and ranking is adopted in the evolution process, and filial generations are generated through crossover and variation; then, by updating ideal points, calculating convergence indexes and distribution indexes and combining angular point solutions to screen individuals, a next-generation population is constructed; and continuously iterating until ending, and finally obtaining a non-dominated solution set as a recommended configuration scheme for graphic rendering performance optimization. The method is suitable for various rendering engines and application scenes, and the operation stability and the user experience of the mobile terminal graphic rendering system can be remarkably improved.

The embodiment of the invention relates to the field of new energy battery lamination processes, and discloses a lamination model parameter adaptive optimization method, device and equipment and a storage medium, and the method comprises the steps: randomly generating a plurality of particles in a preset parameter space according to a lamination model of a battery lamination machine, and respectively representing a group of lamination model parameters; according to the lamination operation process executed by each group of lamination model parameters, the fitness value of each group of corresponding lamination model parameters is obtained according to the offset of the swing rod of the battery lamination machine; defining the distance and direction of each particle during each iteration movement, obtaining updated lamination model parameters during each iteration, and calculating an updated fitness value; and judging whether a preset convergence condition is met or not according to the updated fitness value, if so, outputting a corresponding lamination model parameter, and if not, continuing iteration. The problems that the iteration time is long and the optimization precision is low due to the fact that polling is carried out by adopting a fixed step length in an existing lamination model parameter optimization method can be solved.

The application provides a mechanical arm trajectory optimization method based on model memory data mapping and related equipment, and relates to the technical field of robot control. The application uses a visual language model to quickly identify and position the object to be worked on in the current work scene, quickly understand the expected work task issued by the user for the target mechanical arm, and then combine the model memory data mapping technology, the relationship key point constraint technology and the diffusion strategy technology in the mechanical arm trajectory optimization process. The expected mechanical arm trajectory required for the target mechanical arm to execute the expected work task on the object to be worked on in the current work scene is quickly and accurately optimized, so as to effectively improve the trajectory optimization precision, reduce the trajectory optimization time consumption and the trajectory optimization difficulty, improve the trajectory optimization real-time performance, achieve the synchronous improvement effect of the trajectory optimization real-time performance and the trajectory optimization precision, and ensure that the corresponding mechanical arm can more flexibly adapt to different work requirements of different objects.

The application belongs to the technical field of aircraft engine oil supply and aircraft tank liquid sloshing, and particularly relates to a spatial position optimization layout method of an aircraft engine oil suction port, steps S1: obtaining the maximum mass of fuel consumed by the engine per unit time, determining the cross-sectional area characteristic parameter of the engine oil suction port, determining the minimum distance between the oil suction port and the nearby structure wall plate, and obtaining a first feasible domain in which the oil suction port and the nearby structure wall plate maintain a minimum distance or above; step S2: obtaining multiple working conditions that affect the fuel movement of the tank, obtaining the limit position of the oil surface trajectory under each working condition through simulation, and obtaining the common part of the limit positions of the oil surface trajectory under all working conditions as a second feasible domain; step S3: obtaining the common part of the first feasible domain and the second feasible domain as the final selection area of the best position of the engine oil suction port, and when the first feasible domain and the second feasible domain have no common part, modifying the tank parameters and returning to step 1.