32results about How to "Avoid local extrema" patented technology

The invention provides a power distribution network fault positioning method based on improvement of a binary particle swarm algorithm, the conventional binary particle swarm algorithm is improved, and the method is applied to positioning of power distribution network faults. The method comprises following steps: firstly, determining parameters including the particle swarm scale and the maximum iteration frequency etc.; then forming an expectation function of a switch according to fault information of the switch, and constructing a fitness function of power distribution network fault positioning; initializing a particle swarm, setting particle positions, and setting the speed of the particles as 0; calculating the fitness values of the particles according to the fitness function, and setting an initial global extremum; updating an individual extremum and the initial global extremum; updating the speed and position of the particle swarm; and stopping calculation when reaching the maximum iteration frequency, and outputting the global optimal position of the particle swarm, namely the practical fault state of each feed line section of a target power distribution network. According to the method, the problem of premature convergence of the conventional method can be overcome, and the convergence and the stability of the algorithm can be further improved.

The invention provides an Internet picture filtering method and device. The Internet picture filtering method comprises the following steps: according to different preset dimensions, zooming an obtained Internet image to generate pictures of different dimensions; calculating the quick characteristics of the pictures of different dimensions; adopting preset image annotation information and the quick characteristics to train to generate a target image classifier; utilizing the target image classifier to detect the Internet image, and determining a candidate image area which contains target attributes; inputting the candidate image area into a convolutional neural network to calculate the characteristics of the convolutional neural network; and utilizing the classifier to classify whether the picture contains a target image or not according to the convolutional neural network to realize picture filtering. Internet pictures can be favorably filtered, calculation efficiency is improved, a user can carry out fine tuning on a deep neural network on line, and therefore, detection performance of the Internet picture filtering method is better than the detection performance of other methods.

The invention discloses a hydro governor parameter optimization method based on the fuzzy self-adaptive DFPSO algorithm. The method comprises the steps of (1), establishing a hydroturbine regulating system mathematical model; (2), determining a fitness function of the fuzzy self-adaptive DFPSO algorithm; (3), implementing fuzziness setting on the speed inertia factor of the algorithm, wherein the linearly-decreasing inertia factor and the current particle optimal performance evaluation index are input fuzzily; (4), calculating the fitness values of particles, maintaining the individual optimal values and the global optimal values of the particles and updating the speeds and positions of the particles; (5), carrying out gene crossover between the particles when the crossover condition is met; (6), judging whether the end condition is met, stopping gene crossover and outputting the optimal value when the end condition is met, and implementing the steps (4) to (6) when the end condition is not met. According to the invention, a system can have good dynamic property on the conditions of frequency disturbance and loading disturbance, small overshoot, short stabilization time and short adjustment time.

The invention discloses a particle swarm optimization based orthogonal wavelet blind equalization method. The method comprises the following steps of: allowing a transmitted signal a(k) to pass through a pulse response channel h(k) to acquire a channel output signal x(k); acquiring an orthogonal wavelet transformation (WT) input signal y(k) through channel noise n(k) and x(k); performing WT on the input signal y(k) to acquire an output signal R(k); taking the input signal y(k) as input of a particle swarm optimization (PSO) algorithm and randomly initializing a group of weight vectors, wherein each particle corresponds to each group of weight vectors one to one; determining a fitness function of PSO through a cost function of an orthogonal wavelet transformation-constant module algorithm (WT-CMA) blind equalization method; when a fitness value is the maximum, finding out an optimal position vector in the group and taking the optimal position vector as an initialization weight vector W(k) of the WT-CMA; and acquiring an equalizer output signal z(k) from the output signal R(k) and initialization weight vector W(k). In the method, the optimal equalizer initialization weight vector is sought through PSO, and the autocorrelation of the signal is reduced by WT. Compared with WT-CMA, the method has higher convergence rate and lower steady-state error.

The invention provides a support vector machine method based on chaos and grey wolf optimization, specifically,the grey wolf algorithm chaotization is combined with the support vector machine; two key parameters of penalty coefficient C and kernel width Upsilon of the support vector machine are optimized by the grey wolf chaotization algorithm with outstanding whole searching ability; the best extreme parameter value of the learning machine is obtained so that the application can obtain accurate and intelligent decision effect; the machine can help the decision mechanism effectively to make scientific decision, so the machine has important application value.

The invention relates to a wind power variable-pitch multi-variable fuzzy neural network PID control method. The control method includes the following steps that a fuzzy parameter setting module is used for presetting the weight of a PID neural network module; the error between a rotating speed reference value and the actual rotating speed output of a wind driven generator is calculated through a PID calculation module to obtain a reference output quantity of torque of the wind driven generator; the error between the power output value and the power reference value of the wind driven generator and the error change rate are set through the fuzzy parameter setting module to obtain a presetting parameter of the weight of the PID neural network module; through a negative gradient algorithm with a momentum factor, the weight of the PID neural network module is trained, and the reference value output of torque and the reference value output of the pitch angle of the wind driven generator are adjusted. The output power of the wind driven generator can be stabilized nearby a rated valve, and safety of a fan is ensured.

The invention discloses a four-gear air spring with an adjustable volume and a method for controlling the four-gear air spring. The four-gear air spring with the adjustable volume comprises a closed air bag body, a first auxiliary air chamber and a second auxiliary air chamber. The top of the air bag body is communicated with a main air tube, the lower side of the air bag body is communicated with the first auxiliary air chamber via a first through hole and is communicated with the second auxiliary air chamber via a second through hole, a main electromagnetic valve is mounted on the main air tube, a first electromagnetic valve is mounted on the first through hole, and a second electromagnetic valve is mounted on the second through hole. When the air spring is located at a static balance operational altitude, the volume of the second auxiliary air chamber is identical to that of the air bag body, and the volume of the first auxiliary air chamber is the double of that of the air bag body. The method for controlling the four-gear air spring mainly includes steps of measuring mechanical property curves; building and verifying models; optimizing programs; and executing the programs and the like.

An intelligent energy storage system grid-connected real-time control method based on an artificial fish swarm algorithm adopts an Elman neural network to predict intraday power load real time data on the base of power load historical data, then utilizes the artificial fish swarm algorithm to plan the optimal charge-discharge time and the optimal power of intraday power load prediction data, and performs comparison with the electric power real time data through an intelligent electric meter, so as to determine the optimal charge-discharge time node. The invention achieves automatic grid-connected discharge in the peak of power utilization and achieves charge in the low ebb of power utilization, achieves peak load shifting on the user side, and improves the utilization efficiency of electric power resources.

The invention discloses a method for improving a salp swarm algorithm, and aims to improve the salp swarm algorithm to overcome the defects that the salp swarm algorithm cannot perform accurate searchin the later stage of iteration, is poor in population diversity and the like. By adding an attenuation factor, the search range is flexibly controlled and the algorithm convergence speed is increased, and by introducing a dynamic learning strategy, the assistance effect of a follower on optimization is enhanced, higher convergence precision of the algorithm is achieved, and the optimization performance of the salp swarm algorithm is improved. The convergence precision and the convergence speed of the improved salp swarm algorithm are greatly improved.

The invention provides a method for converting a seabed sonar image into an acoustic substrate classification based on a wavelet neutral network. In the method, an algorithm of a genetic wavelet neutral network is utilized to perform local analyzing; network initial parameters are optimized through a genetic algorithm, so as to avoid trapping in small local, and effectively avoiding noise and local extreme value, and the conversion between the seabed sonar image and the acoustic substrate classification is more precise and reliable, thus, the method provided by the invention has significant practical value in seabed substrate classification.

The invention discloses a control method and system for realizing polarization stability, and belongs to the technical field of optical fiber communication and optical fiber sensing. According to thecontrol method and system of the invention, the output light polarization state of any input light polarization state after passing through a polarization controller is rapidly adjusted to be close toany specified target polarization state on a Poincare sphere through a rapid positioning algorithm, and then the output light polarization state is further stabilized to a target polarization state through a random gradient descent algorithm. Therefore, any polarization state can be stabilized to any set target polarization state. The specific implementation system is composed of an input end polarization analyzer, an output end polarization analyzer, a calibrated polarization controller, a single-chip microcomputer and the like; the input polarization analyzer and the polarization controllerwith the calibrated phase difference and voltage relation are used for rapid positioning, and meanwhile, the output polarization analyzer is combined with the stochastic gradient descent algorithm tofinally stabilize the output polarization state to a set value. According to the polarization stability control method, the searching speed and the local extreme value avoidance are balanced, and thepolarization state can be rapidly and stably changed to any designated target polarization state.

The invention discloses a blast furnace ironmaking multi-objective intelligent optimization method based on an adaptive genetic algorithm. According to the self-adaptive genetic algorithm, the population fitness skewness coefficient is continuously calculated in the iteration process. The population scale is automatically updated according to the change trend of the population fitness skewness coefficient so as to obtain the optimal search performance. The method is applied to blast furnace ironmaking process index multi-objective optimization. Aiming at different crude fuel qualities, production conditions and market conditions, a factory has different requirements on various indexes of the blast furnace. The fitness function of the genetic algorithm is solved by setting the weight of each index through the furnace length. The population size is automatically updated according to the positive and negative change trend of the fitness function in the evolution process so as to ensure that the algorithm has the optimal optimization performance. By applying the self-adaptive genetic algorithm to the ironmaking process, the problem of multi-target optimization of mutual coupling of blast furnaces can be effectively solved. Compared with a traditional optimization algorithm, the method has the advantages that local extremum can be effectively avoided, and a globally optimal solutioncan be efficiently and accurately solved.

The invention relates to a method for improving ant colony algorithm optimization support vector machine parameters. The method includes the steps that the value range of n parameters is determined, and each parameter is equally divided into N parts to calculate the grid interval; ants select N grid points from the first row to the Nth row, the travel of the N grid points serves as one solution, and M ants find out M solutions; the M solutions are input into an objective function, and a largest objective function value and a smallest objective function value are found out; global information element updating is performed, Pt=Pt-1*rho, a certain number of information element values are added within a certain range near to the globally optimal solution according to the formula Pt=Pt-1-op, and the globally optimal solution is strengthened; a certain number of information element values are reduced within a certain range near to the globally worst solution according to the formula Pt=Pt-1-wp, and the globally worst solution is weakened; if the globally largest cycle index is not reached, the grids are redistricted again until the loop termination conditions are met and optimization of the parameters is completed. The method improves the speed and the accuracy rate for searching for the optimal combination. The principle of grids and high-probability random selection is fused in the method, and the sensitiveness of the ants on the optimal solution is increased.

The invention provides a prediction model method based on an improved moth optimization algorithm, which comprises the steps of loading a data set and performing standardized processing on sample data; Gaussian mutation strategy and chaotic agitation are used to improve the moth flame optimization algorithm, and support vector machine model and / or limit learning machine model are constructed by using the improved moth optimization algorithm. The implementation of the invention can not only increase the diversity of the population and enhance the searching ability of the algorithm, but also prevent the algorithm from falling into the local optimum and quickly finding the global optimum solution.

The invention discloses a dynamic contrast enhancement magnetic resonance image detection method, by which non-rigid registration measured based on complexity and similarity of a residual is applied to detection of a magnetic resonance image. The dynamic contrast enhancement magnetic resonance image detection method comprises the following steps: preprocessing an image, performing image interpolation by a B-spline function method, calculating the measuring similarity, setting circulating iteration of a threshold, and working out a difference image as a relatively precise image area display image. The invention as MRI contrast dynamic study, and a conventional pixel intensity values based on a similar approach as compared to the non-rigid registration, under the same experimental conditions, with more accurate registration of the resulting image. In the dynamic contrast enhancement magnetic resonance image detection method, a dynamic contrast magnetic resonance image is taken as a study object; and compared with the conventional non-rigid registration method based on similar intensity values of pixels, the dynamic contrast enhancement magnetic resonance image detection method has the advantage that a image registration result is more accurate under the same experimental condition.

The invention belongs to the field of image processing, and provides a non-rigid medical image registration method and system. The non-rigid medical image registration method comprises the steps of extracting feature points of a reference image and a floating image, and performing coarse registration on the floating image to obtain an initial position of the floating image; taking the feature points corresponding to the reference image and the floating image as control points of a B spline energy field, and carrying out fine registration on the floating image by adopting a local region multi-level non-uniform B spline; calculating the similarity measure of the reference image and the floating image after registration, judging whether the similarity measure meets a preset similarity measureor not, and if yes, outputting registration parameters; otherwise, adjusting the multi-level non-uniform B-spline parameters of the local area to continue to perform fine registration on the floatingimage until the preset similarity measure is met. According to the invention, a local region multi-layer non-uniform B spline is adopted, and grids are further refined for a region with a poor registration effect, so that the registration precision and efficiency are improved.

The invention discloses a denoising method for a seismic data signal based on a variational principle. The method includes the following steps: S1. inputting a test signal f and a reference signal g in a seismic signal processing model and selecting a sampling point; S2. performing normalization processing on the test signal and the reference signal to construct a difference matrix; S3. convertingthe difference matrix to a slowness matrix; S4. based on a Fermat principle and Huygens-Fresnel principle, extracting global minimum travel time and a corresponding shortest path in the slowness matrix by utilizing a minimum path ray tracing method; and S5. applying time shift amount of the travel time of each of sampling points in the obtained shortest path to the test signal to obtain the denoised seismic data signal. The method is modified from two aspects of a theory and implementation, the computational efficiency is improved, the stability of finding the shortest path is enhanced, and anew perspective is opened for measurement of nonlinear signals.

The invention discloses a novel sliding-mode prediction fault-tolerant control algorithm for a discrete uncertain multi-time-lag four-rotor system under an actuator fault. For the fault-tolerant control problem of the discrete uncertain multi-time-lag four-rotor system under the condition that the actuator fault exists, firstly, a quasi-integral sliding mode surface is designed to serve as a prediction model to eliminate an approaching mode, so that the global robustness is guaranteed; secondly, aiming at the actuator fault and multiple time lags, an improved fault compensation double-power function reference track is designed, so that the influence of the time lags on the system is weakened, and the fault-tolerant control precision is improved; and thirdly, an improved inverse time limitcoyote optimization algorithm (ICOA) is designed for rolling optimization, so that while a good convergence rate is obtained, the situation that local extremum is caught in the optimization process isavoided, and the local development and global search performances are balanced. The fault-tolerant control algorithm is used for robust fault-tolerant control of the multi-time-lag discrete uncertainsystem with the actuator fault.

The invention discloses a hydro governor parameter optimization method based on the fuzzy self-adaptive DFPSO algorithm. The method comprises the steps of (1), establishing a hydroturbine regulating system mathematical model; (2), determining a fitness function of the fuzzy self-adaptive DFPSO algorithm; (3), implementing fuzziness setting on the speed inertia factor of the algorithm, wherein the linearly-decreasing inertia factor and the current particle optimal performance evaluation index are input fuzzily; (4), calculating the fitness values of particles, maintaining the individual optimal values and the global optimal values of the particles and updating the speeds and positions of the particles; (5), carrying out gene crossover between the particles when the crossover condition is met; (6), judging whether the end condition is met, stopping gene crossover and outputting the optimal value when the end condition is met, and implementing the steps (4) to (6) when the end condition is not met. According to the invention, a system can have good dynamic property on the conditions of frequency disturbance and loading disturbance, small overshoot, short stabilization time and short adjustment time.

The invention discloses a method and a system for suppressing the crosstalk noise of the full waveform inversion seismic source code. The method includes: performing low-pass filtering on the velocity model updated in the inversion iteration process; performing principal component analysis on the velocity model after the low-pass filtering to determine the main signal direction of the stratum geological structure; performing directional filtering along the main signal direction to suppress Source coded crosstalk noise in a direction different from the main signal. The invention utilizes the direction difference between the signal and the noise in the full waveform inversion speed model, and filters in the direction of the effective signal, which can not only protect the integrity of the effective signal, but also remove the noise.

The invention discloses a seismic phase feature recognition waveform inversion method based on a progressive data assimilation method, which includes dividing seismic wave time windows. Compare the waveform similarity of data in each window, and filter qualified waveforms for waveform inversion. After each iteration, according to the updated forward modeling data of the model, the waveform similarity between the observed data and the theoretical data in each time window is re-compared, and the qualified waveforms are selected for the waveform inversion of the next iteration. The invention divides the exploration seismic data into time windows according to a fixed length, and compares the waveform similarity to screen the seismic data for waveform inversion to solve the problem of cycle jump and improve its convergence efficiency. The invention compares the waveform data in each time window Screening helps solve the cycle jump problem of waveform inversion and improves the convergence of waveform inversion.

The invention discloses a crowd evacuation simulation method and a crowd evacuation simulation device, which can provide evacuation path guidance for crowds, prevent falling into local optimum and improve evacuation efficiency. The crowd evacuation simulation method comprises the following steps: obtaining a crowd image, and preprocessing the crowd image; Extracting texture features of the crowd image; Estimating and classifying the crowd density by utilizing an extreme learning machine algorithm according to the texture features of the crowd image to obtain a crowd density level; and according to the crowd density grade, simulating the crowd image by utilizing an artificial fish swarm algorithm to obtain an optimal evacuation path.

The invention discloses a control method and system for realizing polarization stabilization, and belongs to the technical field of optical fiber communication and optical fiber sensing. In the present invention, the polarization state of the output light after the polarization state of any input light passes through the polarization controller is quickly adjusted to the vicinity of the polarization state of any specified target on the Bonga sphere through the fast positioning algorithm, and the polarization state of the output light is further stabilized to the target by the stochastic gradient descent algorithm. polarization state to achieve stabilization of any polarization state to an arbitrarily set target polarization state. The specific implementation system is composed of two polarization analyzers at the input and output ends, a calibrated polarization controller, and a single-chip microcomputer. Combined with the stochastic gradient descent algorithm, the output polarization state is finally stabilized to the set value. The polarization stabilization control method of the present invention takes into account both the search speed and the avoidance of local extreme values, and realizes the rapid and stable change of the polarization state to any specified target polarization state.

The invention discloses a novel sliding mode predictive fault-tolerant control algorithm for discrete uncertain multi-time-delay four-rotor systems under actuator faults. Aiming at the fault-tolerant control problem of discrete uncertain multi-delay quadrotor system in the presence of actuator faults, a quasi-integral sliding surface is designed as a predictive model to eliminate approaching modes and ensure global robustness. Secondly, aiming at actuator faults and multiple time delays, an improved fault compensation double power function reference trajectory is designed to weaken the influence of time delays on the system and improve the accuracy of fault-tolerant control. Thirdly, an improved Inverse Coyote Algorithm (ICOA) is designed for rolling optimization, which can avoid falling into local extremum during the optimization process while obtaining good convergence speed, and balances the performance of local exploitation and global search. The invention is used for the robust fault-tolerant control of a class of multi-time-delay discrete uncertain systems with actuator faults.