112results about How to "Improve optimization accuracy" patented technology

The invention provides a wind turbine generator system yaw system control performance optimization method and system. The method comprises the steps of acquiring the data of incoming flow air in front of a machine cabin at fixed time intervals, wherein the data of incoming flow air includes the incoming flow air speed, the incoming flow absolute wind direction and the wind direction of incoming flow relative to the machine cabin; sectoring the incoming flow absolute wind direction according to the preset angle intervals, and sectioning the incoming flow air speed according to the preset wind speed intervals; carrying out the first time of grouping on the data of the incoming flow air according to sectors and subsections; calculating the yaw errors of the corresponding groups according to the wind direction of the incoming flow relative to the machine cabin to obtain a yaw error optimization model; and when the yaw system is optimized, inputting the measured incoming flow air speed and the measured incoming flow absolute wind direction into the yaw error optimization model, carry outing matching to find the corresponding yaw error, and correcting the corresponding yaw error to the input of a yaw control system. With the wind turbine generator system yaw system control performance optimization method and system, different optimizing strategies can be adopted for wind turbine generator systems of different models in the situations with different wind speeds of incoming flow, and the optimizing accuracy of yaw errors is improved.

The invention discloses an intelligent pipeline arrangement optimization method and system based on a differential evolution algorithm. The method includes the steps of conducting mathematical modeling on a pipeline to be arranged and arrangement space so as to determine an arrangement object, the constraint condition and the evaluation criteria, coding the arrangement object through polar coordinates and posture vectors, conducting optimization solution on a pipeline arrangement optimization mathematical model according to the differential evolution algorithm, and conducting constraint condition verification and arrangement adjustment on the arrangement optimization scheme obtained through the solution so as to obtain a final arrangement scheme. By means of the intelligent pipeline arrangement optimization method and system based on the differential evolution algorithm, the design cycle can be greatly shortened, optimization performance can be enhanced, the purpose of arranging pipelines on a large scale within limited time can be achieved, and the method and the system have the advantages of being short in arrangement design time, high in optimization accuracy, capable of quantitatively evaluating the arrangement scheme and the like.

The invention provides an LSSVM (least squares support vector machine) wind speed forecasting method based on integration of GA (genetic algorithm) and PSO (particle swarm optimization). The method comprises the following steps: finite wind speed samples are divided into a training set and a testing set, and normalization processing is performed; GA and LSSVM related parameters are initialized; chromosome coding is performed, and initial population is generated randomly; the fitness corresponding to each chromosome is calculated, if requirements are met, the PSO in the fifth step is started directly, and if the requirements are not met, selection, crossover and mutation operation of the GA are performed; optimum parameter combination obtained with the GA is used for initializing the PSO related parameters; the optimum position fitness value of each particle is compared with the optimum position fitness value of the swarm; the final optimum parameter combination is output, and an optimized LSSVM model is obtained; a forecast wind speed time history spectrum is obtained. The LSSVM wind speed forecasting method based on integration of GA and PSO has the characteristics of high optimization precision, high convergence precision, fewer iterations, high success rate and the like.

The invention discloses a driving motor system performance evaluation method for an electric vehicle. The driving motor system performance evaluation method analyzes from different dimensions such as motor control performance, motor body design and enterprise qualification and ability of the driving motor system according to performance characteristics of a driving motor used for the electric vehicle, adopts an analytic hierarchy process to determine a driving motor performance evaluation index system and index weights thereof, establishes a BP neural network model for driving motor system performance evaluation, organically integrates a bat algorithm with a particle swarm algorithm to form a bat-particle particle swarm hybrid algorithm, and optimizes parameters of the neural network structural model by adopting the bat-particle particle swarm hybrid algorithm. Simulation examples show that, through training and testing data samples, the driving motor system performance evaluation method which optimizes the neural network based on the analytic hierarchy process and the bat-particle particle swarm hybrid algorithm has the advantages of fast evaluation speed and high accuracy rate, achieves satisfying evaluation results, and has certain promotion value in evaluation, selection and application of a driving motor system for the electric vehicle.

The invention provides a vehicle spare part sales volume forecasting method and system based on a unified dynamic integration model and a meta-heuristic algorithm. The method comprises the steps thata database is established to store data needed for forecasting the vehicle spare part sales volume, and the sales volume of various vehicle spare parts is comprised and is called as a forecasting variable; a data acquisition module is connected with the database and the vehicle spare part sales volume forecasting system to acquire the needed forecasting variable, and a number of parallel typical forecasting methods are used for forecasting to acquire forecasting results corresponding to various forecasting methods; furthermore, various forecasting results are stored, and a unified dynamic integrated model is established; the meta-heuristic algorithm is used to optimize the forecasting model coefficients; the acquired forecasting model is stored in a vehicle spare part sales volume forecasting application system; and a spare part sales volume forecasting result is generated after the corresponding vehicle spare part sales volume data are input. According to the invention, the model which is suitable for forecasting various vehicle spare parts is found; the characteristics of high optimization precision and the like of the meta-heuristic algorithm are used; and the vehicle spare partsales volume forecasting precision is effectively improved.

The invention relates to a method for memory estimation and configuration optimization in a distributed data processing system. The method at least comprises the steps that a data program stream analyzed and processed for a conditional branch and / or a loop body of a program code in an application jar package is matched with a data feature library, based on a result of successful matching, the memory upper limit of at least one stage is estimated, based on the memory upper limit, configuration parameter optimization is carried out on an application program, based on the running processes of the optimized application program, the static feature and / or dynamic feature of program data are collected, and persistence recording is carried out. According to the method for the memory estimation and configuration optimization in the distributed data processing system, different from a black box model of memory estimation by machine leaning, the accuracy of the result of machine leaning prediction is not necessary high, and fine-grained prediction at each stage is made difficultly; program analysis and existing data features are used to accurately estimate the overall memory footprint, thememory use situation of job at each stage is estimated according to the program analysis, and the further fine-grained configuration optimization is made.

The invention provides an LSSVM fluctuating wind speed prediction method based on integration of an ant colony and particle swarm, and the method comprises the following steps: performing normalizing treatment; calculating pheromone concentration of each ant; moving other ants to the position of the head ant in the ant colony to perform global search; in an iteration process, updating the ant pheromone concentration at each position, inspecting whether an iteration terminal condition is met or not, if not, returning to step 3; otherwise, ending an algorithm and outputting an optimal parameter combination; initializing related parameters of the particle swarm; comparing the fitness value of a self optimal position of each particle with the fitness value of the optimal position of the swarm; obtaining the predicted fluctuating wind speed time interval spectrum. The method has the characteristics of high optimized speed, high convergence precision, less iteration times, high successful rate and the like.

According to the running speed optimization method for the urban rail transit, the energy-saving speed curve is short in calculation time and small in error through the coasting search method, and theoptimization precision of the running speed of an urban rail transit train is effectively improved; according to the method, optimization of the speed curve and the inter-station operation time is comprehensively considered, under the condition that the total operation time is not changed, the situation of sub-interval merging is considered, and the multi-station traction energy consumption is further reduced; according to the method, the adjustability of the speed is optimized for the train speed curve; the coasting starting point advancing step length can be self-defined, the advancing steplength is increased within the time error allowable range, the solving speed can be increased, and through the design, on the basis of safety and punctuality, rapid solving of the energy-saving trainoperation working condition and the working condition switching point can be achieved.

The invention discloses a valve flow function optimization method based on unit operation data. According to the method, operation data of a thermal power generating unit is processed to obtain actual valve flow performance data of the thermal power generating unit; finally, an ideal valve flow function is established on basis of the actual valve flow performance data; configuration of a digital electro-hydraulic system of the thermal power generating unit is set with the ideal valve flow function so that an actual valve flow function is consistent to the ideal valve flow function, and the valve flow function is optimized. Compared with the prior art, the method has the advantages that the valve flow function can be optimized without performing a valve flow performance test, optimization time is shortened, and optimization efficiency is improved; the influence of main steam pressure is fully considered in the calculation of theoretical steam flow, a calculation error caused by fluctuation of the main steam pressure is avoided, and optimization precision is improved.

The invention provides an operation optimization method of a complex parallel cascaded pump station system. The method comprises the following steps of: firstly, building a related optimization model, supposing that a dual-line cascaded pump station system comprises n cascaded levels of pump stations, connecting two water supply lines between the k level pump station (k = 1, 2,..., n) and the k+1 level pump station in parallel to mix the two water supply lines, and then, supplying water to a destination in two paths; and after the constraint conditions of water level, flow, pump unit head of delivery and number of pump stations to be turned on are satisfied, dispersing the to-be-optimized parameters in upper two layers by aiming at minimizing the total input power of the system pump stations and using a multilayer decomposition-dispersion method, namely, solving an in-station optimization scheme of the lower two layers by using a simulated annealing particle swarm optimization algorithm, calculating the total operation power of the whole parallel cascaded pump station system, and considering the in-station optimization scheme as the optimized operation scheme of the whole engineering system if the total operation power is lowest. The method provided by the invention can be used for operation optimization of large parallel cascaded pump stations, expectedly, can save energy by 3%-8%, makes full use of the efficiency of the pump stations, promotes establishment of harmonious society and has great social and economic benefits.

To detect the performance degradation of an optical communication transmission system after it becoming operational, thereby preventing the reliability of the system from lowering. In an optical communication transmission system, one of optical transmitter-receivers in a redundant-structured system is an active optical transmitter-receiver, and the other optical transmitter-receiver is a standby optical transmitter-receiver. The optical transmitter-receiver includes a loopback device through which the signal light of a transmitting section of the optical transmitter-receivers is inputted to a receiving section, and an operating-system switching scheduling device for regularly switching the operating system is provided. Further, there is also provided a performance check scheduling device which operates the loopback device of the standby-system optical transmitter-receiver at a prescribed timing determined in advance to start the performance check processing by the transmitting section and the receiving section of the optical transmitter-receiver.

The invention discloses a method for predicting a membrane pollution tendency in a membrane distilled water processing system on the basis of a GA-LSSVM (Genetic Algorithm- Least Squares Support Vector Machine) model. The method comprises the following steps: firstly, utilizing an LSSVM algorithm to establish a prediction model in a membrane distilled sewage processing process; secondly, utilizing a GA to optimize the parameter of the prediction model independently under a quasi-steady state and an unsteady state; thirdly, utilizing the optimized prediction model to predict the change tendency of membrane flux and membrane pollution resistance independently under a quasi-steady state and an unsteady state, and analyzing an influence on the membrane flux and the membrane pollution resistance by the basic operation parameter of membrane distilling; and finally, carrying out sensitivity analysis and calculation on a prediction result, and determining a leading factor which affects the membrane flux and the membrane pollution resistance. The method utilizes GA-LSSVM to predict a change situation of the membrane flux and the membrane pollution resistance in real time, and the influence on membrane pollution by the basic operation parameter of the membrane distilling is clarified and quantized.

The invention provides an economic optimization method of energy storage cost and benefit, which comprises the following steps: obtaining energy storage investment cost and benefit information; The obtained energy storage investment cost and income information are brought into the preset energy storage optimization objective function and constraint conditions in the whole life cycle by adopting particle swarm optimization algorithm, and the calculation is carried out to obtain the best economic benefit of the energy storage in the whole life cycle of the photovoltaic application; The optimization objective function and constraint conditions are constructed based on the cost and benefit of the energy storage investment. The technical proposal provided by the invention aims at maximizing thetotal net income in the whole life cycle of the energy storage system, analyzes the income and the expenditure of the energy storage system in the photovoltaic power grid, realizes the maximization of the whole life cycle economic benefit of the energy storage application in the photovoltaic power generation based on the investment cost and the income of the energy storage.

The invention discloses a soft measurement method of reactant concentration in a continuous stirred tank reactor based on positive-strand RNA-GA and meter thereof, which is used for solving the problem of variable on-line estimation that on-line real-time measurement is difficultly carried out by using a sensor or analysis meter in the process of a continuous stirred tank reactor. The on-site collected input / output data in various working conditions is used as sample data of RBF neural network after treatment; the Centroid and the weight of the RBF neural network are optimized by using genetic algorithm based on positive-strand RNA so as to obtain the optimal neural network soft measurement model; and the on-line real-time measurement of reactant concentration in a continuous stirred tank reactor is realized by using the soft measurement model. The invention has high precision and real-time performance, and provides necessary conditions for efficiently optimizing and controlling the continuous stirred tank reactor in advanced manner.

The invention provides an involute straight tooth gear modification optimization method; and the method comprises the steps of: building engaged three-dimensional gear pair models; primarily selecting modification optimization parameters; respectively performing the primary modification processing for the three-dimensional gear pair models; performing the primary modification optimization analysis to obtain optimal primary analysis optimization parameters; obtaining more precise optimization samples in finite element analysis software according to the optimization parameters; and screening the samples. After screening, the secondary modification processing is performed for a pair of standard three-dimensional involute gears to obtain multiple sets of secondarily processed optimization parameters, optimal modification optimization gear parameters, one set of parabola modification and one set of arc modification; the three-dimensional diagram drawing is performed for gears optimized by the secondarily processed optimization parameters; and optimally modified three-dimensional diagrams are leaded in a numerical control machine tool to machine blanks to finished products. The method is more precise in modification, is more obvious in optimization effect, can preferably reduce the engagement noise and vibration, and reduces such adverse effects as load concentration.

The invention relates to the technical field of urban sewage treatment, and particularly discloses a multi-objective optimization method for an urban sewage treatment process under multiple working conditions, which comprises the following steps: 1) constructing mathematical description of a multi-objective optimization problem of the urban sewage treatment process; 2) designing a multi-objectiveparticle swarm optimization algorithm in the urban sewage treatment process; and 3) designing a multi-working-condition case library in the urban sewage treatment process. The method aims at the characteristic that the uncertainty is high under different working conditions in the sewage treatment process; a real-time dynamic optimization multi-target particle swarm algorithm is designed, meanwhile, the concept of a multi-working-condition case library is introduced, for matched cases, the corresponding optimal solution in the cases can be directly applied to population initialization of the current working condition, and the search precision and convergence speed of the algorithm are improved. The concentration set values of dissolved oxygen and nitrate nitrogen in the urban sewage treatment process under multiple working conditions are optimized in real time, and energy consumption is effectively reduced on the basis that the effluent quality reaches the standard.

The present invention discloses an optimum design method for an anterior angle and a relief angle of a groove milling tool, and belongs to the technical field of mechanical manufacturing. The method is characterized by: by comparing and analyzing a temperature value of the tool under different combination conditions of the anterior angle and the relief angle, obtaining an optimal anterior angle and an optimal relief angle of the tool. A theoretical calculation of the tool temperature and an orthogonal experiment design method are combined by the method of the invention to optimize the anterior angle and the relief angle of the tool, and an optimize accuracy is higher. According to the present invention, the method is free from a cutting experiment, so that an optimization period and an optimization cost is greatly reduced. The method of the present invention can be widely applied in a groove milling condition, and in various matching phenamena of the tool and a workpiece material.

The invention discloses a method for optimizing controller parameters of an automatic power grid generation control system, relating to the field of automatic control of an electric power system, aiming at solving the technical problems that the controller parameters are not proper so that a dynamic control performance of the automatic power grid generation control system is relatively low. The method for optimizing the controller parameters comprises: establishing an automatic power grid generation control system simulation model, wherein the automatic power grid generation control system simulation model comprises a controller; according to the automatic power grid generation control system simulation model, establishing an optimized modulation model of the controller parameters; according to the optimized modulation model, obtaining an initial optimization result of the controller parameters through an ecological niche bacterial foraging algorithm; and according to the initial optimization result, obtaining a final optimization result of the controller parameters through a pattern search algorithm. The method provided by the invention is applied to optimization of the controller parameters of the automatic power grid generation control system.

The invention provides a prediction method of an LSSVM non-Gaussian fluctuating wind speed, which comprises seven steps, wherein the specific steps are that a non-Gaussian random fluctuating wind speed sample is generated through simulation with a memory-less nonlinear conversion method, the non-Gaussian fluctuating wind speed sample is divided into a training set and a testing set, and normalization is carried out to the two sets respectively; and training learning is carried out to the LSSVM by the training set, the testing set is used for prediction, fitness of each chromosome in a groups is calculated, whether algorithm convergence criterions are satisfied, a combined solution is put into a set if an optimal parameter combination is satisfied and then the fifth step is started, and otherwise the fourth step will be started. The prediction method of the LSSVM non-Gaussian fluctuating wind speed provided by the invention combines a genetic algorithm and an ant colony algorithm to intelligently extract the optimal parameter combination of the LSSVM in order to establish an optimized LSSVM prediction model and predict the testing set. In this way, a predicted time interval spectrum of the non-Gaussian fluctuating wind speed can be obtained.

The invention discloses a pure electric vehicle frame lightweight method based on nonlinear programming. The method comprises the following steps of S1, establishing a finite element model of a frame;S2, determining an optimized objective function and an optimized design variable, and solving a mathematical expression of the objective function by means of the Hypermesh software; S3, determining an optimized constraint condition, and collecting the sample data of each constraint response; S4, processing the sample data, performing function fitting on each constraint response by using a second-order response surface model, and solving an approximate agent model of each constraint; and S5, constructing a nonlinear programming model by using the agent model of each constraint and the mathematical expression of the objective function, and solving the optimization model. According to the method, by adopting the nonlinear programming model, and using the Lingo software for solving and optimizing, the optimal solution can be obtained within the extremely short time, and the optimization efficiency is greatly improved. The optimization method has the characteristics of simple principle, high solution rate, high optimization precision and the like.

The invention discloses a sparse limited non-negative matrix decomposition algorithm based ultrafiltration membrane water treatment prediction method. Firstly, a sparse limited non-negative matrix decomposition algorithm is utilized to construct a prediction model for a membrane distillation wastewater treatment process; secondly, a GA algorithm is utilized to optimize parameters of the prediction model in a pseudo-steady state and a non-steady state respectively; thirdly, the optimized prediction model is utilized to predict the change trend of the membrane flux and the membrane pollution resistance and to analyze the influence of basic operation parameters of membrane distillation on the membrane flux and the membrane pollution resistance in the pseudo-steady state and the non-steady state respectively; and finally, a prediction result is subjected to sensitivity analysis calculation to determine leading factors which influence the membrane flux and the membrane pollution resistance. According to the method, the sparse limited non-negative matrix decomposition algorithm is utilized to predict the change situations of the membrane flux and the membrane pollution resistance in real time, and the influence of the basic operation parameters of membrane distillation on the membrane flux and the membrane pollution resistance is clarified and quantified.

The invention relates to an electromagnetic mechanism static characteristic optimization method based on an adaptive weight multi-objective differential evolution algorithm, which comprises the following steps: S1, determining parameters to be optimized and static characteristic related indexes of an electromagnetic system; S2, determining the upper limit and the lower limit of each static characteristic optimization parameter according to the product material and the processing technology of the electromagnetic mechanism, and simultaneously determining an additional constraint index related to the static characteristic optimization parameter; S3, obtaining an initial population; S4: obtaining the Pareto solution set distribution of the optimization objective function; S5, selecting different mutation strategies and crossing strategies according to the number of iteration times of the current optimization parameter population and the Pareto dominance relationship of each optimization parameter, generating a progeny population, repeating the operation of step 4 to the whole population, and adopting a selection strategy considering the niche sorting result to control the population size; S6, obtaining the Pareto solution set distribution of the optimization objective function; S7, selecting a group of optimization parameters from the obtained groups of optimization parameters, and taking the group of optimization parameters as the optimization design parameters of the electromagnetic mechanism.

The invention discloses a method for optimizing a multi-span optical fiber transmission system, which comprises the following steps of carrying out abstract analysis on an optical fiber and an amplifier in the multi-span optical fiber transmission system, extracting factors influencing the system performance and modeling; carrying out derivation calculation on the established mathematical model according to different methods under set conditions to obtain an expression for representing the performance of the optical fiber transmission system; extracting the to-be-optimized quantity from the expression of the performance of the optical fiber transmission system, optimizing the to-be-optimized quantity by using the genetic algorithm, and setting the optical fiber transmission system by usingthe optimized parameters, so that the performance of the optical fiber transmission system can be enhanced. According to the method, multi-factor modeling is carried out by abstractly analyzing and extracting influence factors of the multi-span optical fiber transmission system, so that the expression accuracy of the model on the transmission system is improved; a system transmissibility expression is obtained by deriving a mathematical model, and parameters are optimized by utilizing a genetic algorithm, so that the optimization precision is improved compared with a traditional optimizationalgorithm, and the transmission performance of the system is improved.

The invention provides an oil field dynamic intrusion target tracking system and method based on multiple unmanned aerial vehicles, and relates to the technical field of unmanned aerial vehicle inspection and dynamic prediction tracking. The system comprises a control module, a monitoring module, a dynamic target trajectory prediction module and a multi-machine cooperative target tracking trajectory optimization module. According to the system, the monitoring module mounted on the unmanned aerial vehicle is used for collecting data of a dynamic intrusion target in a ground oil field working environment, then the dynamic target trajectory prediction module is used for predicting trajectories of the target in multiple sampling moments in the future, and the multi-machine cooperative target tracking trajectory optimization module is used for optimizing the distance error between the unmanned aerial vehicle and the target in the tracking process according to the predicted trajectories, sothat the distance error is kept within an effective tracking threshold value, and cooperative target tracking of the multiple unmanned aerial vehicles is achieved.

An optimal Latin hypercube design (LHD) method based on an adaptive genetic algorithm relates to a Latin hypercube design method and aims to solve the defects of premature convergence, local optimal solution and high optimal cost produced in a conventional optimization algorithm optimizing process. The method comprises the steps of I, representing an initial population Npop formed by n LHD(N,P) individuals as a k*N matrix; II, calculating adaptation values of the n individuals in the population Npop, and selecting the individual with a maximum adaptation as a current optimal LHD(N,P); III, ranking the n adaptation values from largest to smallest, selecting first n / 2 individuals as parents, and generating n descendants by use of the selected n / 2 parents and forming a new population; IV, carrying out adaptive variation on the new population to obtain a new altered population; and V, iteratively carrying out steps II-IV, and judging whether an iteration termination condition is satisfied. The method is used for the field of computer simulation experimental design.

The invention relates to a networked intelligent vehicle formation and running control method on the basis of spatial domains. The networked intelligent vehicle formation and running control method includes that front vehicle information of controlled vehicles can be acquired by control systems of the controlled vehicles, the controlled vehicles can be communicated with front vehicles to acquire the front vehicle information, head time intervals and head distances of head locations of the controlled vehicles and the front vehicles when the front vehicles run through the current locations of the controlled vehicles can be computed by the control systems, the front vehicle information at the intervals of certain distances on roads can be computed by the control systems at the current head time intervals of the controlled vehicles and the front vehicles, controlled vehicle information and the computed front vehicle information can be combined with each other by the control systems, acceleration, brake and steering procedures of the controlled vehicles can be optimized, optimization results can be transmitted to power devices, brake devices and steering devices of the controlled vehicles, and vehicle running can be optimized by the aid of optimization principles. The networked intelligent vehicle formation and running control method has the advantages that networked intelligent vehicle longitudinal control and transverse control procedures can be simultaneously optimized, the optimization precision can be improved, repeated iteration can be omitted by optimization algorithms, and accordingly computation load can be relieved.

The invention provides a physical cell identifier (PCI) interference self optimization method and a PCI interference self optimization device. The method comprises steps: the correlation between a first PCI and a second PCI is acquired, wherein the first PCI is the PCI corresponding to a target cell, and the second PCI is the PCI corresponding to an adjacent cell with the same frequency as the target cell; based on measurement reports, backstage switching indexes and sweep frequency data of the target cell and the adjacent cell with the same frequency, a PCI comprehensive interference factor of the target cell is determined according to the correlation; if the PCI comprehensive interference factor of the target cell is larger than a threshold value, the target cell is determined to be a PCI high interference cell; and the target PCI is selected for the target cell according to a preset rule, and the PCI corresponding to the target cell is changed to be the target PCI. The embodiment of the invention effectively solves the limitations of the traditional PCI optimization, the PCI optimization accuracy is improved, and the PCI change can be adapted to network changes flexibly and timely.

The invention relates to an RFID network reader scheduling optimization method based on a multi-swarm particle swarm algorithm. The method comprises the following steps: the RFID reader network is initialized; the size of a fitness value of each particle is detected; the speed of a symbiosis swarm is updated, and an individual extreme value, a swarm extreme value and a global extreme value are selected; discrete operation is carried out on the position of the particle; the maximum iteration number is set to be a termination criterion, if a reader still does not complete operation after the iteration of the time, a reader sub graph working in the time slot is deleted in an RFID interference pattern, and the step of detecting the size of the fitness value of each particle is returned; or otherwise, an RFID network reader scheduling result is outputted. Simultaneous evolution and mutual cooperation among multiple swarms can be realized, the solving efficiency is high, the operation is simple, the global search ability is strong, the convergence speed is quick, the optimization precision is high, and a new solution scheme is provided for solving the continuous optimization problem in a practical engineering application.

The invention discloses a fractional order KVFD multi-parameter machine learning optimization method for viscoelasticity mechanical characterization of soft substances. The method comprises the steps:building corresponding K-tree dictionaries according to the conditions of solutions of three types of KVFD loading modes; judging the specific type of a to-be-tested curve, carrying out global search, obtaining the parameter [E0, alpha, tau] of the curve as a vector, zooming to a parameter interval, generating a preset number of curves according to the KVFD model corresponding to the to-be-testedcurve, adding random Gaussian noise, dividing into a training set and a test set, and transmitting the training set and the test set into a machine learning model for training; selecting the model with the minimum RMSE as a final model for training; performing parameter estimation on a to-be-measured curve by using the final model obtained in the step 3; and further learning a result obtained byparameter estimation through a Q-learning algorithm to obtain an optimization result. According to the method, the characteristics of parameter learning and a heuristic algorithm are combined, and theaccuracy and efficiency of parameter optimization can be greatly improved.

The invention provides an econometrics and heuristic intelligence combined automobile sales prediction method and system. The method comprises the following steps of: storing data required by automobile sales prediction through establishing a database, wherein data comprises economic indexes, brand automobile sales and automobile sales, which are called as prediction variables; connecting the database and the automobile sales prediction system through the data so as to obtain required prediction variables, and verifying structural relationships between the variables so as to obtain endogenousvariables with a long-term equilibrium and causal relationship; establishing a vector error correction model for the endogenous variables; optimizing coefficient of a prediction model by utilizing a heuristic intelligence algorithm; and finally storing the obtained prediction model into a sales prediction application system, and generating a sales prediction result after inputting corresponding economic variables, brand automobile sales data and former-year automobile sales data. According to the method and system, a model suitable for long-term prediction is found, and the characteristic of high precision of the heuristic intelligence can be utilized at the same time, so that the automobile sales prediction precision is effectively improved.