56results about How to "Realize dynamic optimization" patented technology

The invention discloses a dynamic optimization method for structural parameters of mechanical equipment parts based on digital twin. The method comprises the following steps: realizing digital mirrorimages of equipment corresponding to physical space in virtual space by constructing a high-fidelity model, which is convenient for later structural parameters modification and super realistic simulation; dynamically and synchronously reflecting the state of an entity corresponding to the physical space in real time in the virtual space by performing super realistic simulation, and realizing realistic motion of the real condition of the physical equipment in the virtual space; and meanwhile, constructing a neural network structure by utilizing a deep learning theory, establishing a relation between the structural parameters and the fatigue life with the help of strong digital mining and mapping capabilities of the neural network structure, and realizing the dynamic optimization of the structural parameters by combining the high-fidelity model and the super realistic simulation environment. According to the invention, dynamic optimization and reverse guidance of the virtual space to thestructural parameters of the physical space are realized, and optimization efficiency and authenticity are improved.

The invention relates to a production method of an oil deposit, in particular to an oil deposit development production optimization method and system based on a connectivity method. The method comprises the steps that saturation calculation is conducted by taking inter-shaft communication units as objects to establish a precise front edge tracking method to obtain oil-water dual-phase production dynamics of each layer at shafts; information of inter-shaft conductivity, flow splitting, water injection efficiency and the like is obtained by inverting connectivity model parameters though automatically fitting oil deposit historical dynamics, and the information is taken as the basis to perform layered dynamic production allocation and injection allocation automatic optimizing design on the oil deposit through iterative computation to reduce low-efficiency water drive direction flow and improve injection-exploitation conflict. Application examples show that the method can perform rapid computation, accurate saturation tracking and production dynamics computation, an obtained layered water injection scheme can well achieve a water reducing and oil increasing effect, 50,000 cubic metersof predicted oil increasing after actual optimization of the oil deposit is conducted is achieved, and a novel idea for performing real-time optimization on an oil field injection-exploitation schemeis provided.

The invention discloses a magnetic suspension rotor harmonic current inhibition method based on parallel FORC and phase lag-lead compensation, and the method comprises the steps: firstly building a magnetic suspension rotor dynamical model comprising quality imbalance and sensor harmonic; secondly employing a magnetic suspension rotor harmonic current inhibition method based on a parallel FORC structure. The structure enables internal models of a harmonic at all frequencies to be converted into a plurality of parallel units with the similar structures, and quickly improves the dynamic performances of a system. The method also employs the phase lag-lead compensation to improve the stability and dynamic performances of the system. The invention proposes a method for quickly and adaptively adjusting the controller gain for a specific harmonic signal. The method achieves the dynamic optimization of the dynamic performances and stability. The method achieves the inhibition of a harmonic component of a current of a magnetic bearing coil in a magnetic suspension rotor, and is suitable for the harmonic current inhibition of the magnetic suspension rotor system with the quality imbalance and sensor harmonic.

The invention belongs to the fields of engineering machinery and vehicle engineering, and particularly relates to a shift strategy dynamic-optimization method based on deep reinforcement learning. Theshift strategy dynamic-optimization method based on deep reinforcement learning comprises the following steps that (1), the state input variable and action output variable of a shift strategy are determined; (2), the Markov decision process of the shift strategy is determined according to the state input variable and the action output variable; (3), according to the shift strategy objective, a reward function of the reinforcement learning shift strategy is established; (4), the deep reinforcement learning shift strategy is solved according to the Markov decision process and the reward function; (5) a prediction Q network calculated in the step (4) is put in a shift strategy controller, during the driving process of engineering machinery and vehicles, the engineering machinery and the vehicles select gears according to the shift strategy controller; and (6), the prediction Q network is regularly updated in the driving process. The shift strategy is updated through a deep reinforcementlearning method to realize the dynamic optimization of the shift strategy.

One or more embodiments of the invention disclose an information prediction method and device based on label propagation, which are used for realizing high efficiency and high accuracy of informationprediction based on a label propagation mode. The method comprises the following steps: acquiring first tags of specified types of a plurality of sample users; acquiring behavior information of each sample user and a to-be-predicted user; determining an initial label of the to-be-predicted user according to the first label and the behavior information; determining a dynamic label propagation algorithm corresponding to each iteration; and, according to the first label, iterating the initial label by using the dynamic label propagation algorithm, and when the iteration meets a preset convergencecondition, determining that the label obtained by the last iteration is the second label of the specified type of the user to be predicted.

The invention discloses a single road intersection traffic signal control method based on real-time traffic information. The single road intersection traffic signal control method comprises the steps of: establishing a single road intersection traffic signal safety control mathematical model based on real-time traffic information; solving the mathematical model by adopting a self-adaptive timing algorithm based on real-time traffic information; setting green-light release shortest time, green-light release longest time and green-light release time in saturated traffic flow respectively through terms and marks, decision variable and objective functions; judging whether a safety accident occurs in multiple lanes; and judging the queuing length of a current phase position. On the basis of establishing the single road intersection traffic signal safety control mathematical model based on real-time traffic information, the single road intersection traffic signal control method based on real-time traffic information adopts the self-adaptive timing algorithm for controlling the intersection according to real-time vehicle queuing length and traffic safety state, and achieves dynamic optimization timing of single road intersection traffic signals on the premise of ensuring traffic safety.

The invention discloses a synthesis aperture heat antenna feeder system suitable for large posture changes. The synthesis aperture heat antenna feeder system is composed of two synthesis aperture heat antenna feeder antennas and an antenna controller. The synthesis aperture heat antenna feeder antennas are installed on an air vehicle in a vertically-opposite mode. The multi-band frequency conformal cold-heat-integrated phased array design mode is adopted for the synthesis aperture heat antenna feeder antennas, multi-band frequency antennas are embedded into phased-array array faces in a layered mode, and transmitting and receiving of Ka / Ku frequency band signals, S / C frequency band signals and L frequency band signals can be completed through the single synthesis aperture heat antenna feeder antenna. The phased-array array faces and antenna heat protection covers of the synthesis aperture heat antenna feeder antennas are integrally designed in a synthesis mode, and the amplitudes and the phase positions of the antennas can be corrected by adjusting phased-array antenna parameters. The antenna controller obtains the angle and the distance between the air vehicle and an object according to position information and posture information of the air vehicle, and the number of working transmitting array elements and the number of working receiving array elements of the synthesis aperture heat antenna feeder antennas are dynamically adjusted; meanwhile, the synthesis aperture heat antenna feeder antennas suitable for communication are selected according to the angle information, and beam control is completed.

The invention discloses an on-line time series prediction method based on a dynamic fuzzy cognitive map, and the method includes the following steps: forming information granules and mapping a time series to an activation level of the information granules to form nodes of a fuzzy cognitive map; optimizing the information granules on a parameter level of the fuzzy cognitive map, and determining theweights of the nodes of the fuzzy cognitive map; and dynamically adjusting the fuzzy cognitive map according to data coming in at a current moment, and utilizing the dynamically adjusted fuzzy cognitive map to perform time series prediction. The on-line time series prediction method based on a dynamic fuzzy cognitive map adjusts a clustering center and parameters according to the specific situation, realizes dynamic optimization, and enables a model to capture the change of data information in real time, thereby enabling the prediction to be more accurate.

A knowledge and data information decision making-based sewage processing optimization control method is disclosed and belongs to the field of intelligent control. According to the method, decision making on weight of multi-target optimization indexes is performed via fusion of knowledge information and data information; sewage processing procedure optimization performance indexes can be improved while that effluent quality can reach standards is be ensured. In a first step, weight coefficient distribution of the multi-target optimization indexes is determined according to sewage processing knowledge information and process dynamic data information used for decision making; in a second step, a group of Pareto optimal solutions for optimization problems are be obtained via a multi-target optimization method, and a satisfactory optimization solution, i.e. optimization set values for dissolved oxygen concentration and nitrate nitrogen concentration, is determined according to obtained weight coefficients; in a final step, an optimization set value tracking task is carried out via a bottom layer controller; according to the method, knowledge information on a subjective layer and dynamicdata information in processing procedures are taken into consideration during decision making direction adjustment, and energy consumption and operation cost in sewage processing processes can be effectively lowered.

The invention provides a modified loofah seed oil-based memory sponge. The modified loofah seed oil-based memory sponge is prepared from the following raw materials in parts by weight: 1-40 parts of loofah seed oil-based polyol, 40-60 parts of first polyether polyol, 10-30 parts of second polyether polyol, 10-20 parts of third polyether polyol, 1-5 parts of water, 5-15 parts of a foaming agent, 0.5-3 parts of a foam stabilizer, 0.3-2 parts of a catalyst, 0.1-0.3 part of an antioxidant and 40-60 parts of modified MDI. The invention also provides a preparation method and application of the modified loofah seed oil-based memory sponge. Meanwhile, the invention further provides the modified loofah seed oil memory sponge pillow. The modified loofah seed oil-based memory sponge material is prepared by mixing loofah seed oil-based polyol with polyether polyol, modified MDI and the like through a polyurethane slow-rebound foaming technology. The sponge is used as a pillow inner material and has the advantages of being antibacterial, capable of helping sleep, breathable, capable of dissipating moisture, good in mechanical performance, low in cost, environmentally friendly and the like.

The invention discloses a method for optimizing the rolling scope of hot strip rolling. The method comprises the following step: setting initial parameters; selecting the width node value and initial thickness value of a rolled piece; generating an initial rolling schedule of the rolled piece according to the width node value and initial thickness value of the rolled piece; judging whether rolling constraint conditions are satisfied or not; reducing the thickness of the rolled piece according to the thickness variable step size of the rolled piece; recording and outputting the last computed minimum thickness value of the rolled piece, i.e. the corresponding minimum thickness value of the width node; and calculating the next width node value. The method aims at relative equal load and good sheet shape, takes finish rolling temperature, allowable rolling force and allowable power as inequality constraints, utilizes pressing distribution as equality constraint to optimize a pressing schedule, accordingly, the minimum rolling thickness is searched, the optimum formulation of a rolling technology is realized at the same time, the equipment capability can be utilized more effectively, the yield of a rolling machine and the product quality are improved, and the rolling energy consumption and production cost are reduced.

The invention discloses a respiratory muscle training system in a motion state and a using method thereof. The respiratory muscle training system comprises a wearable respiratory monitoring module, aninteractive display module, a sports apparatus and a respiratory training management system. The wearable respiration monitoring module monitors parameters such as a respiration rate, a respiration flow rate, a respiration flow and heart rate of a subject in real time through a flow sensor and an electrocardio-electrode and transmits the parameters to the interactive display module. The interactive display module receives a training mode and training parameters issued by the training management system, compares the training mode and training parameters with respiratory and heart rate information obtained from the wearable respiratory monitoring module, and reminds a subject to adjust the respiratory rhythm and respiratory depth in real time; the sports apparatus is used for assisting a subject to keep respiratory training under set exercise intensity, so that the training effect can be improved; the respiratory training management system is a management tool for a rehabilitation doctor to sign and issue a training scheme and is used for making a respiratory training plan, setting differentiated training parameters, tracking and managing respiratory training process data and realizing continuous optimization of the respiratory training scheme.

The invention provides a server platform power consumption optimizing method and device. The method comprises: reading configuration type of a system; acquiring a correspondence table for corresponding CPU (central processing unit) usage and power consumption according to the configuration type of the system; reading a current stable value of CPU usage; setting an upper limit of power consumptionaccording to the value of the CPU usage; adjusting operating frequency of a CPU according to the upper limit of power consumption; monitoring whether the value of CPU usage is stable; if not, releasing the upper limit of power consumption that is already set; after the value of CPU usage is stable, resetting the upper limit of power consumption corresponding to the current value of CPU usage. Theserver platform power consumption optimizing method and device have the advantages that the upper limit of power consumption can be adjusted in real time at the premise of ensuring server performance,the power consumption of a server is dynamically optimized, and power consumption cost is reduced for the server.

Embodiments of the invention relate to an unmanned ship global meteorological route dynamic planning method and system. The method comprises the following steps of acquiring an initial route and discretizing the initial route to obtain a plurality of initial waypoints, wherein the multiple initial waypoints include a starting point, an ending point and a plurality of intermediate waypoints locatedbetween the starting point and the ending point; collecting real-time meteorological data in the ship driving process; for the current waypoint of ship driving, performing multilevel decision in combination with the real-time meteorological data of the multiple waypoints from the initial waypoint to the current waypoint by position decision of the initial waypoint to obtain multiple correspondingoptimal waypoints; and dynamically planning non-navigation routes after the current waypoint according to the multiple optimal waypoints and the ending point. According to the method, real-time dynamic adjustment and planning are carried out on the positions of the waypoints on the initial route in a segmented multilevel decision mode, and dynamic optimization of the routes in the ship driving isrealized.

The embodiment of the invention discloses a data processing method, apparatus and device, and a readable storage medium. The method comprises the steps of obtaining video frame data and audio frame data of a target video; determining position attribute information of a target object in the target video according to the video frame data, wherein the target object is associated with the audio framedata; acquiring a sound channel coding parameter associated with the position attribute information, and performing orientation enhancement processing on the audio frame data according to the sound channel coding parameter to obtain enhanced audio frame data. By adopting the method and the device, the audio optimization effect in the video can be improved.

The invention relates to the technical field of network monitoring, and discloses a network request safety verification method, a device, a medium and electronic equipment. The method comprises the steps of receiving a first network request, generating a first token according to the first network request, and sending the first token to a sending end of the first network request, wherein the firsttoken comprises failure time; determining the time for monitoring the trigger operation of the sending end of the first network request; when the trigger operation of the sending end of the first network request is monitored at the time, adjusting the failure time of the first token according to the trigger operation to obtain the adjusted failure time of the first token; and when a second networkrequest carrying the first token is received before the adjusted failure time, confirming that the second network request is legal. According to the method, after the token is generated for the network request, the failure time of the token is adjusted according to the triggering operation of the user, so that the dynamic optimization of the failure time of the token is realized, and the safety of the token is considered while the user experience is ensured.

The invention belongs to the field of data centers. The invention discloses a data center PUE dynamic optimization method, system and device and a readable storage medium. The method comprises the following steps: obtaining operation data information of each server in the data center, determining an operation state of each server according to the operation data information, obtaining a plurality of to-be-migrated calculation tasks, respectively scheduling the plurality of to-be-migrated calculation tasks to each server in the low-load operation state according to the operation data informationuntil all servers in the low-load operation state reach a full-load operation state. According to the method, the computing task to be migrated is scheduled to the server in the low-load running state to reach the full-load running state so that the problem of large deviation between the actual PUE value and the theoretical PUE value under the condition of long-term running due to the non-uniformload of the server is solved, and the PUE value of the data center is effectively reduced.

A network-connected hybrid electric vehicle energy management system of a distributed MPC comprises an engine, a clutch, a motor, a transmission and batteries. The MCU of the motor is connected with the TCU of the transmission and the ECU of the engine; the batteries are respectively connected with the motors; the batteries transmit power to the motor; working modes comprise pure electric driving, pure engine driving, hybrid driving, driving charging and regenerative braking; the method by means of the network-connected hybrid electric vehicle energy management system comprises the following steps: 1, modeling a parallel hybrid electric vehicle; 2, predicting the speed of the networked hybrid electric vehicle; and 3, constructing a distributed MPC energy management method; the system and the method have the characteristics of good calculation efficiency, fuel economy and working condition adaptability.

The invention discloses a method for realizing dynamic hydraulic optimization of a central heating pipe network system, which is based on a nonlinear programming algorithm and comprises the followingsteps: 1) calibrating pipe network resistance, heat user resistance, an electric valve characteristic curve and a water pump characteristic curve of the centralized heating pipe network system, wherein the pipe network and the heat user resistance are calibrated by combining a pipe network design drawing, a pressure gauge and a heat meter actual measurement parameter; according to the electric valve characteristic curve, under a certain water supply pressure, a fitting characteristic curve is obtained by measuring a plurality of flow and valve opening corresponding values; according to the water pump characteristic curve, pipe network resistance is adjusted by closing an electric valve, and a fitting characteristic curve is obtained by measuring a plurality of total water supply and returnflows and pressure differences; 2) establishing a hydraulic equilibrium equation; and 3) taking minimum water pump conveying energy consumption as a target, and taking electric valve resistance, thermal inlet flow and total water supply and return pressure difference as constraint variables.

The invention discloses a sidewalk pedestrian and vehicle self-sensing early warning system based on a piezoelectric pavement. The sidewalk pedestrian and vehicle self-sensing early warning system comprises a comprehensive indicating rod, a piezoelectric speed measuring device, a piezoelectric vehicle sensing device, a piezoelectric pedestrian sensing device and a sidewalk piezoelectric sensing device. The piezoelectric speed measuring device is arranged on the road surface on one side of the sidewalk, and is used for sensing the time difference before and after the vehicle runs by utilizing the piezoelectric road surface so as to obtain the running speed of the vehicle. The piezoelectric vehicle sensing device is installed in front of a red light vehicle stop line and used for detecting vehicles passing through the line in the red light period. The piezoelectric pedestrian sensing devices are buried in the two sides of a sidewalk and used for dynamically sensing waiting pedestrians. The sidewalk piezoelectric sensing device is arranged on a sidewalk and used for sensing the walking state of pedestrians. The comprehensive indicating rod is installed beside a sidewalk and used for displaying and reminding the state of pedestrians and vehicles. The early warning system can effectively perform early warning, ensures pedestrian safety, and reduces sidewalk traffic accident rate.

The invention relates to an optimization method and device for a query statement, a storage medium and computer equipment. The method comprises the following steps that: obtaining the query statement,and obtaining a query tree according to the query statement; beginning to traverse the query node of the query tree from the root node of the query tree; when the query node which conforms to a preset execution rule is in the presence, obtaining query data corresponding to the query node as current query data; when an association attribute value is in the presence in the current query data, extracting the association attribute value, and pushing the association attribute value to the corresponding query node; when an associated query node which is not optimized is in the presence in the querynode, obtaining the query data of the query node, and judging whether the execution cost of the query node exceeds the preset execution cost or not; and if the execution cost of the query node exceeds the preset execution cost, taking the query data as the current query data to enter the step when the association attribute value is in the presence in the current query data until the associated query node corresponding to the association attribute value does not contain the associated query node which is not optimized. By use of the method, query efficiency during cross-database joint query isimproved.

The invention relates to the technical field of digital marketing, and discloses a marketing activity operation method and device, and the method comprises the steps: inputting the demand information of a marketing activity into a marketing activity recommendation model, and outputting a marketing activity scheme; wherein the marketing activity recommendation model is obtained by training according to platform operation data and historical marketing activity data; according to the marketing activity scheme, determining configuration information of a marketing activity page and issuing a marketing activity; according to the marketing effect data of the marketing activity and the user feedback information, adjusting the configuration information of the marketing activity page and publishing the adjusted marketing activity, the marketing activity recommendation model is obtained based on training of the platform operation data and the historical marketing activity data, various activity factors can be considered, the accuracy of a marketing activity scheme is improved. According to the marketing effect data and the user feedback information, the configuration information of the marketing activity page is adjusted in real time, dynamic optimization of the marketing activity can be realized, and the real-time performance of operation management of the marketing activity is improved.

The invention discloses a dynamic optimization method and system. The method comprises the following steps: obtaining running information of all running tasks and selecting the to-be-optimized tasks with performance bottlenecks from all the tasks based on the running information; optimizing the to-be-optimized tasks by utilizing a pre-optimization principle input from the outside; obtaining optimized running information of the optimized to-be-optimized tasks and returning the corresponding to-be-optimized tasks to states before optimization when the optimized running information does not meet the preset requirements. Through the technical scheme disclosed by the invention, the running information of all the running tasks is obtained, and the to-be-optimized tasks are determined based on the running information; the to-be-optimized tasks are optimized, so that dynamic obtaining of the running information of all the running tasks is realized, dynamic optimization of the tasks is further realized, the effectiveness of task optimization is ensured, and efficient running of the various tasks in the operating system is ensured.

The invention discloses a numerical control machine tool cutting stability prediction and optimization method considering parameter uncertainty, and the prediction method comprises the steps: defining a machine tool as a chain structure product work structure model, and constructing the chain structure product work structure model through a software simulation platform; constraints including a machine tool pose and a main shaft rotating speed are added in the software simulation platform so as to complete construction of a finite element model of the whole machine tool of the generalized dynamic machining space; based on a whole machine finite element model of a generalized dynamic machining space machine tool, influence factors including a machining point frequency response function and a cutting force coefficient are analyzed and calculated through a robust stability prediction model, so that a high-speed cutting variable parameter stability prediction model is established; and stability verification is carried out on the edge theorem in the robust stability prediction model through a zero arrangement principle, a robust high-speed cutting stability lobe graph is obtained, and prediction of the cutting stability is completed. The predicted stability domain boundary is more practical, and the accuracy of the stability domain boundary is obviously improved.

The invention discloses an adaptive following predictive control method, which comprises the steps of monitoring states of an own vehicle and a preceding vehicle in real time, and obtaining a state information value of each control sub-target in a p-step predictive time domain; calculating variances of the control sub-targets in the current control period and the previous control period respectively according to the state information values; comparing the variance of each control sub-target in the current control period with the variance of the control sub-target in the previous control periodto obtain a weight adjustment factor of the control sub-target; and adjusting the weight coefficient of each control sub-target in a quadratic cost function according to the weight adjustment factorso as to realize dynamic optimization of the quadratic cost function, obtain an optimal control vector and control the own vehicle according to the optimal control vector. The adaptive following predictive control method can improve the adaptability of a multi-target predictive controller to complex working conditions and optimize the control quality. The invention further discloses a self-adaptive following predictive control system and device and a computer readable storage medium, which all have the above technical effects.

The embodiment of the invention discloses an image color filling method, a device, equipment and a storage medium, and the method comprises the steps: in response to a color filling triggering operation acting on a target image, determining an initial current color filling block from an image block set corresponding to the target image; when it is detected that the read-back operation is not executed on the current color filling block, reading all pixel points in the current color filling block back to the cache block; performing color filling on the current color filling block based on each pixel point in the cache block corresponding to the current color filling block; and determining a next color filling block according to the color filling result of the current color filling block, taking the next color filling block as a new current color filling block, and returning to execute a read-back detection operation of the current color filling block until it is determined that the nextcolor filling block does not exist in the target image. By utilizing the method, the whole target image is prevented from being read back to the memory before color filling, the memory resource occupation in the color filling process is reduced, the dynamic optimization of the memory during image color filling is realized, the color filling time of image color filling is shortened, and the color filling efficiency is improved.

The invention discloses a power amplifier voltage adjusting method based on machine learning. The method comprises the following steps of constructing a machine learning model formed based on voltage parameter training, obtaining the current telephone traffic, and predicting the current telephone traffic based on a machine learning model to generate a prediction result, automatically adjusting the power amplifier voltage based on a preset power amplifier adjustment algorithm and prediction results, wherein the prediction results comprise a power alignment result, a time delay alignment result and a peak clipping coefficient updating result. Therefore, according to the method disclosed by the invention, the power amplifier voltage can be dynamically adjusted to improve the overall power amplifier efficiency of the circuit by predicting the change of the telephone traffic under the condition that a turn-off mode is not needed, so that the energy consumption of the power amplifier during working is greatly reduced.