54results about How to "Reduce randomness" patented technology

The invention relates to the technical field of fault detection for a wireless communication network, and discloses a fault detection method for the wireless communication network. By collecting data of base station signal strength, network delay, data packet loss rate and user equipment signal quality, the problem of insufficient fault detection caused by a single index is solved. Through multi-dimensional data acquisition, the system can comprehensively evaluate network performance, the fault detection accuracy is improved, and different fault types can be effectively identified especially in a complex fault mode. By storing the data according to the time sequence, the problem of time sequence chaos is avoided, and accurate detection of time-varying faults is ensured. The comprehensive data vector generated by each time of sampling comprises a plurality of network performance indexes, so that the fault detection and positioning capability is improved. Abnormal values are eliminated through data preprocessing, the data quality is ensured, an accurate basis is provided for subsequent analysis, and the reliability of a detection result is ensured.

The invention relates to the technical field of vehicle control, and discloses a vehicle-mounted semantic standardization method, device and equipment, a medium and a product, and the method comprises the steps: obtaining a pre-trained large language model; obtaining a target original vehicle control command; obtaining a format template used for representing the target standard vehicle control command; forming a cue word according to the target original vehicle control command and the format template; and inputting the target original vehicle control command and the cue word into the large language model, and outputting the converted target standard vehicle control command through the large language model. Based on the ability of the large language model to directly understand the text meaning, the original vehicle control command and the standard vehicle control command are directly paired to generate the training sample of the text style, so that the training sample is input into the large language model for training, the threshold of model training is reduced, and the training efficiency and the training accuracy are improved; and high accuracy and high automation degree of vehicle-mounted semantic standardization are considered.

The invention relates to the technical field of photovoltaic inverters, in particular to a photovoltaic inverter controller parameter identification method and system, and the method comprises the steps: building a mathematical model of a photovoltaic inverter controller, and determining controller parameters needing to be identified; the mathematical model is converted into a difference equation through transformation decoupling operation; the input and output quantity of an inverter controller is obtained in real time to serve as the input quantity of an identification system, and the preset performance index minimization serves as the optimal adaptive value estimation of an optimization algorithm; and stopping iteration when the parameters of the controller to be identified reach preset precision, and outputting estimated values of the parameters of the controller. The method has the advantages that the optimal solution is found through the idea of minimizing the residual module value, the randomness of identifying the parameters of the controller is greatly reduced, the accuracy of the identification result is improved, online identification and synchronous identification of all the parameters can be achieved, and step-by-step identification of all the parameters is avoided.

A straight line fitting method for image detection in a complex interference environment comprises the following steps: firstly, carrying out interference point deletion on a feature point set of an environment image, then carrying out multiple times of random sampling, forming a total point set by using associated points through distance operation, obtaining a plurality of straight line models through initial fitting, and carrying out linear fitting on the linear models; and then calculating an interior point proportion in the total point set according to the linear model, finally selecting the total point set with the maximum number of interior points, and continuously reducing a residual threshold in multiple iterative calculations to improve an interior point proportion threshold by dynamically adjusting a feedback mechanism of the residual threshold when the interior points are calculated. In the initial stage of iteration, more points can enter the model to ensure the diversity and robustness of the model, and the threshold value is gradually tightened along with the convergence of the model, so that the interior points conforming to the model are screened out more accurately. And finally, performing secondary sampling on the inner points and performing straight line fitting on an inner point set subjected to secondary sampling again to eliminate potential error inner points and further improve the fitting precision.

The invention discloses a distributed photovoltaic power prediction method and system based on time sequence data reconstruction. The method comprises the following steps: collecting a plurality of single-class distributed photovoltaic power curves; adopting improved photovoltaic power data empirical mode decomposition to decompose the plurality of single-class distributed photovoltaic power curves into sub-power curves with different frequencies; the sub-power curves of different frequencies are reconstructed; a neural network is adopted to predict the reconstructed sub-power curves, and then superposition and summation are carried out on predicted values of the sub-power curves to obtain distributed photovoltaic power data of a certain type; according to the method, photovoltaic power is classified into single-class curves of different scenes according to different seasons and weathers, and sub-power curve division is performed by using improved empirical mode decomposition, so that the randomness and volatility of original data can be reduced to the greatest extent before prediction, a neural network can be conveniently and efficiently trained, and the prediction efficiency is improved. And outputting accurate photovoltaic power time sequence prediction data.

The invention discloses an old block three-dimensional form intelligent updating optimization method based on rigid-flexible rule driving. The method comprises the steps of collecting vector data of a target old block and design specifications and standard text data of a city where the target old block is located; establishing a rule map of urban design rigid industry specification requirements; a Mask R-CNN coding technology is used to automatically label vector data of a space structure, a function layout, a road system and a blue-green system of the design, a knowledge graph of a classic city design result rule is formed, and a rule graph oriented by city design flexible professional knowledge is established; establishing multi-layer network association of an old block three-dimensional morphological graph network, a rigid rule graph and a flexible knowledge graph; and outputting the optimized three-dimensional form scheme of the target old block to a space holographic sand table, and finally outputting an optimized three-dimensional form graph network and a visualization scheme of the old block. According to the method, the three-dimensional form of the old block can be optimized based on rigid rule and flexible knowledge hybrid driving.

The invention discloses a regional load prediction method, system, equipment and medium, and particularly relates to the technical field of power load prediction, and the technical key points are as follows: a plurality of load nodes are clustered in sequence according to weather data, population flow data and urban power consumption characteristics by using a third-order clustering analysis method to obtain a tree-type clustering cluster topological structure; a representative load node is selected from each cluster subset, average correlation coefficients between influence features in weather data of each representative load node and load data are calculated, and the influence features of the first three maximum average correlation coefficients are selected as influence input features of the representative load node; performing modal decomposition on the load data representing the load nodes by using an EMD decomposition method to obtain a plurality of connotation modal components and residual terms; and inputting the influence input feature of each representative load node and the plurality of decomposed connotation mode components into a load prediction model to obtain a load prediction result corresponding to each representative load node.

The invention discloses an intelligent passenger car obstacle avoidance path planning method based on a genetic algorithm and extended RRT, and belongs to the technical field of intelligent car path planning. Firstly, a sampling flexible boundary is set according to a global path, an RRT algorithm sampling space is reduced, and a feasible path is guided to be close to the global path through a heuristic extension node; designing a metric function for search tree nodes to select father nodes, establishing a passenger car simplified model to detect collision, then constructing a path optimization area, solving an optimal obstacle avoidance path by using a genetic algorithm, and finally smoothing the path by using a Bezier interpolation method; according to the method, the obstacle avoidance local path can be quickly planned in consideration of the path quality and safety, the safety of the vehicle during driving is remarkably improved, and the accident rate of the vehicle is effectively reduced.

The invention belongs to the field of ball ultra-precision machining, and particularly relates to a ball controllable orbital transfer circulating mechanism for a horizontal ball grinding machine and a grinding method. According to the technical scheme, the ball controllable orbital transfer circulating mechanism for the horizontal ball grinding machine comprises a material disc, a ball outlet sliding rail and a ball inlet sliding rail, and the material disc, the ball outlet sliding rail and the ball inlet sliding rail are each of an upper-layer structure and a lower-layer structure; a grinding mechanism of the horizontal ball grinding machine comprises a flat plate and a grinding plate, a material plate is connected with the upper portion of the grinding plate through a ball outlet sliding rail, the material plate is connected with the lower portion of the grinding plate through a ball inlet sliding rail, and a circular conveying mechanism of balls is formed. According to the ball controllable orbital transfer circulating mechanism for the horizontal ball grinding machine and the grinding method, balls can be controllably and alternately ground in the inner grinding orbit and the outer grinding orbit, the grinding area envelope lines of the balls are roughly the same, the sphericity of products of the same batch is more uniform after the balls are machined, and the grinding quality is improved. And the equal cutting probability of the product is improved.

This application provides a method, apparatus, and electronic device for planning an escape route for an automated valet parking vehicle. The method includes: acquiring reference lines, multiple anchor points, and scene type information of the scene where the vehicle is located; calculating the reference line cost of a search point based on the reference lines; calculating the anchor point cost of the search point based on the multiple anchor points; adjusting the weights of the reference line cost and anchor point cost according to the scene type information; and executing a path planning algorithm based on the adjusted reference line cost and anchor point cost to generate an escape route for the automated valet parking vehicle. This method aims to improve the search efficiency of the path planning algorithm.

The present application relates to the technical field of bridge quality detection, especially to a continuous beam bridge different span loading load test method, comprising the following steps: determining a control section on the continuous beam bridge; measuring the axle load and the wheelbase of the test vehicle; calculating and analyzing the interval area of the calculated bending moment change effect function with the moving vehicle; analyzing and calculating the interval area of the measured bending moment change effect function with the moving vehicle; comparing the ratio of the interval area of the calculated bending moment change effect function with the moving vehicle and the interval area of the measured bending moment change effect function with the moving vehicle through the calibration coefficient to determine whether the bearing capacity requirement is met. The continuous beam bridge different span loading load test method can effectively reduce the arrangement of measuring points and the need for test vehicles, and quickly and accurately judge the bearing capacity state of the bridge by calculating the bending moment change effect function with the moving vehicle and the measured bending moment change effect function with the moving vehicle.

The invention discloses a ship aging management object screening method, and the method comprises the following steps: 1), obtaining a ship composition element diagram for a ship product to be subjected to aging management; 2) constructing an aging management demand degree comprehensive evaluation model; 3) obtaining the evaluation grade of each evaluation index of the evaluation object; 4) constructing an evaluation decision evaluation matrix, and performing aging management priority evaluation; 5) determining a solution with the highest priority for elements in the evaluation matrix; 6) calculating the similarity of the cloud model; 7) calculating a correlation coefficient Ri based on cloud model similarity; 8) performing aging management priority evaluation on the m evaluation objects in the ship composition element graph according to the correlation coefficient Ri based on the cloud model similarity; and 9) obtaining the sequence of the aging management priorities of the basic elements in the whole ship range. According to the ship aging management object screening method based on the cloud model theory, fuzziness and randomness generated in the subjective evaluation process are reduced.

A self-adaptive multi-objective evolution method for optimizing the capacity of an off-grid wind and light hydrogen production and energy storage system comprises the following steps: firstly, according to an energy storage system model, taking unit hydrogen production cost, wind and light abandoning penalty and load shedding penalty as minimization objectives; secondly, the capacity of each component of the wind-solar power generation unit, the power storage unit and the hydrogen and oxygen storage unit is used as a decision variable, and the system is set to conform to the actual constraint; thirdly, initializing a population; individuals in the population are initialized according to the constructed search space, so that the individuals meet constraint conditions; then adaptively selecting a search strategy; the strategy adaptive network selects a proper search strategy based on individual decision variables and objective function values, and evolves a new sub-population; and fusing the parent population and the child population to calculate the non-dominated level and the congestion degree of all individuals, determining parameters of the reward updating strategy adaptive network according to the non-dominated level, and screening out a new population. According to the method, the randomness and blindness of algorithm search can be reduced, and the search efficiency of the algorithm is enhanced.

The invention relates to the field of microbiological examination laboratory equipment, and discloses a drying and disinfecting device and method for a specimen glass slide for microbiological examination. The device comprises a machine shell, a motor and a controller, the detachable rotating shaft is in transmission connection with the driving motor; the support frame is detachably arranged on the detachable rotating shaft and forms a plurality of mounting surfaces; the glass slide mounting plate is detachably mounted on the mounting surface, and a plurality of glass slide mounting stations are arranged on the glass slide mounting plate; the drying module comprises a drying lamp; the disinfection module comprises an ultraviolet disinfection lamp; and the control module is electrically connected with the driving motor, the drying module and the disinfection module and is used for executing a drying program and a disinfection program. The problems that in an existing glass slide processing method, the drying capacity and uniformity are difficult to consider at the same time, the drying process is uncontrollable, and low efficiency and safety risks are caused by separation of the drying process and the disinfection process are effectively solved.

This application provides a model training method, a coarse ranking method, an apparatus, an electronic device, and a storage medium. The method includes: inputting model training samples into a coarse ranking model to be trained; encoding user feature data and object feature data of multiple candidate objects to obtain corresponding user feature vectors and object feature vectors corresponding to each candidate object; fusing the user feature vectors and object feature vectors to obtain predicted conversion rate and predicted click-through rate; fusing the user feature vectors with the object feature vectors corresponding to any two candidate objects according to the ranking results to obtain a ranking factor; calculating the loss of the predicted conversion rate, predicted click-through rate, and ranking factor to obtain the loss value of the coarse ranking model to be trained; updating the model parameters of the coarse ranking model to be trained based on the loss value, and iteratively training the coarse ranking model until it converges to obtain the coarse ranking model.

The invention discloses an information-loss-rate-free optical tuning coding artificial visual neuron, a preparation method and a use method. The preparation method comprises the following steps: firstly, growing an oxide layer on a silicon substrate, then forming a photosensitive channel by spin-coating photosensitive quantum dots on a channel layer, then forming an FET source-drain structure through photoetching and thermal evaporation, and then forming a TS parallel structure through atomic layer deposition, photoetching and thermal evaporation, thereby forming a 2TS-FET device and completing the preparation of the artificial visual neurons. The artificial visual neuron tests the peak current of the 2TS-FET device through the peripheral circuit as information coding. According to the invention, the TS parallel connection is utilized to reduce the current refractory period and the channel quantum dots have different sensing capabilities for light with different wavelengths, and the performance of the transistor is shown as the magnitude of channel current, so that the magnitude and frequency of coding current of a device are influenced, and the artificial visual neuron can be optically tuned without information loss. The artificial visual neuron is simple in structure, high in coding precision and low in power consumption and has CMOS compatibility.

The present invention relates to a steam turbine generator rotor end coil forming device, comprising a press forming device and a hydraulic control device, wherein the press forming device is provided with two groups, the press forming device comprises a press forming mechanism and a ring body, the press forming mechanism is provided in a group of two, and multiple groups of press forming mechanisms are evenly distributed on the ring body; the device comprises a hydraulic control mechanism, an oil inlet separator and an oil outlet separator, the oil outlet end of the hydraulic control mechanism is connected to the oil inlet separator, the oil inlet end of the hydraulic control mechanism is connected to the oil outlet separator, the oil inlet separator and the oil outlet separator are arranged on the ring body, the oil inlet separator is connected to the oil inlet end of the press forming mechanism, and the oil outlet end of the press forming mechanism is connected to the oil outlet separator; the hydraulic control device controls the press forming mechanism to press the end of the steam turbine generator rotor coil. The present invention can accurately control the forming of the rotor coil end, and has the characteristics of high precision, stable and reliable system, and high production efficiency.

The invention relates to the technical field of power equipment monitoring, and discloses a short-circuit impedance identification method and system based on VMD noise reduction and probability density optimization, and the method comprises the steps: decomposing an electrical signal through a VMD method, removing random noise, and effectively filtering external interference; meanwhile, the short-circuit impedance is calculated through probability density distribution based on fast neighbor search, so that errors caused by random fluctuation are reduced. According to the method, the reliability of electrical signal acquisition and the accuracy of short-circuit impedance identification can be remarkably improved, the method is suitable for online monitoring of power equipment such as a transformer, and important technical support is provided for equipment fault diagnosis.

The present invention belongs to the technical field of structural optimization of mobile laboratories. The present invention discloses a structural optimization method for a container-modified mobile laboratory based on finite element analysis, comprising: S1, acquiring basic information of the container, performing finite element analysis on the container based on the basic information of the container and obtaining finite element analysis results; S2, dividing the container into regions according to the finite element analysis results, and then obtaining weak areas and non-weak areas of the container; S3, acquiring functional information of the mobile laboratory, determining the number of openings and the positions of the openings in the non-weak areas based on the functional information, and then obtaining n opening strategies; the present invention can divide the weak areas and non-weak areas of the container, and provides a basis for formulating opening strategies; by comprehensively considering the thermal environment performance, airflow distribution performance and stress distribution performance of each opening strategy through an improved genetic algorithm, it is possible to easily select the optimal opening strategy that meets the target performance threshold.

The present invention provides a multi-sequence seabed intelligent water sampling system and method based on deep learning. The system includes a multi-sequence water sampler, an underwater camera, a sonar, an environmental parameter sensor unit and a power unit arranged on the multi-sequence water sampler; and a central processing unit; the central processing unit is used to receive data collected by the underwater camera, the environmental parameter sensor unit and the sonar in real time, and use a pre-trained deep learning model to perform image recognition on the data collected by the underwater camera, and control the multi-sequence water sampler to perform seawater sampling according to the recognition result and the data collected by the environmental parameter sensor unit and the sonar, and control the power unit to drive the multi-sequence water sampler to achieve adaptive movement; the present invention can realize tasks such as accurate selection of seabed sampling points, construction of three-dimensional seabed terrain, adaptive selection of water sampler sampling timing, and adaptive sample detection around seabed sampling points.

This disclosure provides an image processing method, apparatus, electronic device, storage medium, and program product. The method includes: acquiring a first preview image of a target object based on an image acquisition module, and determining the regions where each part of the target object is located in the first preview image; upon detecting a trigger operation targeting a target part among the various parts, determining a first focusing parameter based on the target region where the target part is located; and controlling the camera of the image acquisition module to focus on the target object according to the first focusing parameter. Embodiments of this disclosure can improve the reliability of focusing on target parts of the target object.

The application relates to a wiring mechanism which comprises a mouthpiece, a first structural member and a pressure adjusting assembly, the mouthpiece is used for being connected with a rotating mechanism for winding a wire on a workpiece to be wound, a first clamping part is arranged on the mouthpiece; the first structural member is arranged on the mouthpiece, a second clamping part is arranged on the first structural member, and the first clamping part and the second clamping part are matched for clamping the wire; the pressure adjusting assembly is arranged between the mouthpiece and the first structural member, and the pressure adjusting assembly is used for driving the mouthpiece to move so as to adjust the force applied by the first clamping part to the wire. In the wiring mechanism, the mutual cooperation among the mouthpiece, the first structural member and the pressure adjusting assembly enables the first structural member to apply a set force to the wire, so that the wire has a certain tension, and then the wire can be always hooked by the mouthpiece when wiring, and the wire can be closely arranged on a terminal when winding, thereby reducing the randomness and the defective rate during wiring.

The invention discloses a method and a system for predicting the remaining service life of an energy storage battery. The method comprises the following steps: acquiring battery capacity time sequence data; decomposing the battery capacity time sequence data by adopting TVF-EMD to obtain an intrinsic mode component; optimizing the number of convolution kernels and the size of the convolution kernels of the time convolution neural network model by using an improved collaborative group optimization algorithm; and normalizing the complete battery capacity time sequence data, and inputting the normalized data into the time convolutional neural network model optimized by the improved collaborative group optimization algorithm to predict the remaining service life. According to the method, the time convolutional neural network model is optimized through the improved collaborative group optimization algorithm, so that the prediction has higher accuracy.

The invention relates to a ginseng age lossless classification model construction method and a ginseng age lossless classification system, in particular to a ginseng growth age classification method based on a hyperspectral imaging technology and an improved swarm intelligence optimization algorithm. The method solves the problems of high-dimensional data redundancy, insufficient map feature utilization, insufficient model generalization ability, low precision of fine classification for adjacent years of cultivated ginseng and the like in the prior art. The method comprises the following steps: 1, acquiring hyperspectral image data of ginsengs of different ages and preprocessing the hyperspectral image data; 2, screening spectral features and texture features based on the hyperspectral image data; 3, performing feature-level fusion on the spectral features and the texture features to obtain fusion features; and 4, inputting the fusion features into the constructed IHSO-RF model for training and testing. The classification model is embedded in the system, and classification of the age of the ginseng is achieved. The method is suitable for the fields of intelligent perception and analysis based on hyperspectrum and machine vision, nondestructive testing and grading of quality of agricultural products / traditional Chinese medicinal materials and the like.

This invention relates to the field of testing technology for large ship propulsion systems, specifically to a test method and system for assessing the wear reliability of large water-lubricated bearings. This invention ensures the similarity in lubrication and cooling performance among three pairs of friction pairs tested—material test blocks, scaled-down samples, and full-scale prototypes—in terms of test objects. When comparing and analyzing wear, the influence of relative sliding speed on wear is simplified by using a linear velocity cyclic test design with the same time span, thus simplifying the wear rate model of the water-lubricated bearing. Compared to solely relying on the durability cyclic test data of the full-scale prototype, the addition of scaled-down sample tests reduces the randomness of test data, and the use of the maximum average wear value as input to determine the wear model of the water-lubricated bearing improves the engineering reliability of the assessment method. This invention can shorten the testing and assessment cycle for the wear reliability of large water-lubricated bearings, save testing costs, and provide highly accurate assessment results.

The utility model discloses a semiconductor structure, and belongs to the technical field of semiconductors. The semiconductor structure comprises a first test piece which comprises a plurality of first comb tooth structures; the second test piece comprises a plurality of second comb tooth structures, and the second comb tooth structures and the first comb tooth structures are arranged in a staggered mode; and the test hole structures are arranged on the first comb tooth structure and the second comb tooth structure in a whole row. According to the semiconductor structure provided by the utility model, the detection accuracy can be improved, and the test efficiency of the chip can be accelerated.

The present application provides a vehicle fuel quantity monitoring method, system, computer equipment and storage medium, which relates to the field of vehicle safety technology, wherein the vehicle-mounted sensor collects vehicle fuel quantity data; the cloud server constructs a plane rectangular coordinate system to generate a vehicle fuel quantity change diagram, calculates the image similarity between the fuel quantity change diagrams of each vehicle, determines the reference vehicle with the most similar fuel quantity change situation, and corrects the observed fuel quantity change diagram of the target vehicle according to the reference vehicle to obtain optimized fuel quantity data; the vehicle-mounted processor marks the current fuel quantity status of the target vehicle according to the optimized fuel quantity data, and determines whether the fuel quantity alarm condition is met; if so, determines the fuel quantity alarm level; the vehicle-mounted alarm sends an alarm to the personnel in the vehicle according to the alarm strategy; the user terminal sends an alarm to the remote personnel according to the alarm strategy. The above method is adopted to monitor and manage the vehicle fuel quantity completely and comprehensively, and at the same time improve the effectiveness and timeliness of vehicle fuel quantity monitoring.

This invention relates to a triangular prism-type tuned mass damper suitable for large-span cantilever structures, comprising a triangular prism box (1), a particle damping unit (4) disposed within the triangular prism box (1), and a steel mass block (3). A high-damping spring (2) is located below the triangular prism box (1). The particle damping unit (4) includes a particle damping chamber (5), three different sized particle balls, and a foamed rubber pad (9). When the structural vibration amplitude is small, the particle balls dissipate energy by compressing the foamed rubber pad and rolling; when the structural vibration amplitude is large, the particle balls dissipate energy by colliding and compressing the foamed rubber. Compared with the prior art, this invention achieves a phased energy dissipation mechanism while solving the problems of narrow vibration reduction bandwidth, large motion stroke, and noise generated by particle collisions in traditional mass tuned dampers.

This application relates to the field of vibration equipment technology and discloses a concrete vibrator for core molds in hollow core slab construction. The vibrator includes a vibrating rod and a sliding plate that can move along the upper surface of a reinforcing cage. A sleeve through which the vibrating rod passes is located at the center of the sliding plate. A cross-hinged rocker arm assembly is provided on the sleeve, and each of the four ends of the rocker arm assembly has clamping legs that can contact the reinforcing cage. This concrete vibrator for core molds in hollow core slab construction, through the coordinated design of a movable sliding plate, a mechanical centering clamping mechanism, and a vibrating rod guiding and stabilizing sleeve, transforms vibration operations from a high-risk manual operation relying on individual worker experience into a streamlined, mechanically assisted, accurate positioning and stable operation. This significantly reduces the probability of core mold breakage, improves the consistency of vibration quality, reduces worker labor intensity, and enhances construction safety.

The invention provides an AIE fluorescent probe-oriented multi-index weighted scoring screening method and system. The method comprises the following steps: acquiring data such as a probe molecular structure and an experimental environment, and preprocessing the data to obtain a feature data set; for thermal stability, light stability, Stokes shift and AIE enhancement factors, optimal molecular descriptors and dichotomy prediction models are selected respectively; inputting a to-be-evaluated probe into the target model to obtain a prediction result, and calculating a synthetic accessibility score; combining the prediction result with the standardized synthetic accessibility score to form index data; the weight of each index is determined through mixed data factor analysis, and a comprehensive score is calculated to realize probe sorting and screening. According to the method, efficient screening of the AIE probe can be realized, the trial and error cost is reduced, and the scoring stability and applicability are enhanced.