36results about How to "Suitable for online applications" patented technology

The invention relates to an inner temperature estimation method for a single cell, which comprises an offline part and an online part. The offline part executes the steps of a1) acquiring the internal resistance characteristic of the single cell at different temperatures; and a2) acquiring a standard relationship between the inner temperature of the cell and the internal resistance of the cell according to the internal resistance characteristic. The online part executes the steps of b1) performing online detection on the current, end voltage and surface temperature of a current working cell; be) performing online estimation for the internal resistance of the current working cell according to the current and the end voltage; and b3) estimating the internal temperature of the current working cell according to the internal resistance acquired in the step b2), the surface temperature acquired in the step b1) and the standard relationship acquired in the step a2). Compared with the prior art, the inner temperature estimation method for the single cell has advantages of effective online estimation for the inner temperature of the single cell, simple operation, effective calculation amount reduction, etc.

The invention relates to a rolling procedure making method of an aluminium single-stand cold-rolling machine, in rolling specification obtained by a rolling specification library according to the rolling material and specification, four system parameters are included to be suitable for the characteristics of rolling process of an aluminium single stand, and annealing, thickness, tensile force andspeed parameters in each pass can be respectively determined; the rolling specification is formed by a pass programming system, a thickness distribution system, a speed distribution system and a tensile distribution system, the specification of the rolling material carries out rolling procedure making by obtaining the rolling specification by the rolling specification library, so as to realize automatic generation of the rolling procedures of different materials with different width and thickness specifications in the production process of the aluminium single-stand cold-rolling machine and carry out rolling procedure making. The method has strong practicality, and can automatically realize rolling procedure making for different materials and specifications; the method is simple and practical, has high computing efficiency and rapid speed and is suitable for on-line usage.

The invention discloses a hybrid power system energy management strategy based on reverse deep reinforcement learning. The strategy comprises the following steps of calculating a globally optimized SOC result as expert knowledge by using an optimization solution method; creating a reward neural network; learning the expert knowledge by utilizing reverse reinforcement learning to obtain parametersof the reward neural network; creating an action neural network, and evaluating the neural network; setting an SOC value before vehicle interaction; inputting the obtained SOC value before interactioninto a reward neural network to obtain a reward value; inputting the obtained pre-interaction SOC value into an action neural network to obtain a mode distribution ratio; interacting with the environment by using the mode distribution ratio to obtain an SOC value after interaction; inputting the SOC value before interaction, the mode distribution ratio, the reward value and the SOC value after interaction into an evaluation neural network to obtain an evaluation value; and calculating the gradient of each network and performing back propagation to update the network parameters by an intelligent agent till training is finished. The method is advantaged in that the optimal reward function can be learned from the expert knowledge, so the deep reinforcement learning effect is better.

The invention relates to a method for estimating the internal temperature of a single battery, which includes an offline part and an online part. The offline part includes: a1) obtaining the internal resistance characteristics of the battery cell at different temperatures; a2) according to the internal resistance characteristics Obtain the standard relationship between the internal temperature of the battery and the internal resistance of the battery; the online part includes: b1) online detection of the current, terminal voltage and surface temperature of the current working battery; b2) online estimation of the current working battery according to the current and terminal voltage Internal resistance; b3) Estimate the internal temperature of the current working battery according to the internal resistance of the battery obtained in step b2), the surface temperature obtained in step b1) and the standard relationship obtained in step a2). Compared with the prior art, the present invention has the advantages of real-time estimation of the internal temperature of the battery, simple method, and effectively reduced calculation amount.

The invention discloses a low-voltage power distribution network cascading failure early warning method based on a risk assessment model, and the method comprises the steps: data collection and processing: collecting load data, weather data and fault data of a low-voltage power distribution network, and carrying out the data preprocessing of the collected data; fault feature selection: adopting an improved G-ReliefF algorithm to carry out optimal selection of fault features of the low-voltage power distribution network; performing cascading failure search: performing subsequent failure search according to the line outage model and the key line model, generating an accident chain and evaluating the accident chain; and fault early warning: calculating a risk assessment coefficient of the element and the line, determining a fault occurrence probability through the risk assessment coefficient, finally carrying out comprehensive risk assessment on the fault, and calculating a risk level. The method is used for predicting and mastering the safe operation state of the low-voltage power distribution network, discovering potential dangers and judging the fault development trend so as to improve the accuracy of fault early warning and further avoid large-area power failure.

The invention relates to an online identification and visualization method for power grid disturbances based on multi-resolution wavelet analysis, which belongs to the technical field of wide-area dynamic monitoring and visualization of power systems. The method comprises the following steps: carrying out offline simulation analysis on the features of typical dynamic disturbance accidents in power grid, and selecting a wavelet function Psi and the decomposition layer-number I corresponding to the obtained maximum wavelet coefficient energy as online wavelet multi-resolution analysis tools; carrying out regionalization according to the PMU (power management unit) configuration situation of a target power grid, so that each region has one and only one PMU substation for measurement, and transmitting frequency signals measured by each PMU substation to a WAMS (wide area measurement system) master station; acquiring frequency signals of each PMU substation from the WAMS master station, and carrying out I-layer online wavelet multi-resolution analysis on the frequency signals by using the determined wavelet function so as to obtain wavelet coefficients of an I(th) layer; and discriminating disturbance occurrence time, disturbance occurrence potion and disturbance type according to the wavelet coefficients; and carrying out visualized display through taking a map as the background. The method disclosed by the invention can improve the running reliability of an electrical power grid.

The invention discloses a method for controlling a rolling mode of a pinch pass mill set so as to control 1# and 2# frames of the pinch pass mill. The method can realize the optimum control for specific elongation by choosing a proper rolling mode according to device parameters of a singe and double dual purpose pinch pass mill set and the technique parameters of belt materials, wherein the device parameters comprise a frame stiffness and ultimate tension and the technique parameters comprise the roughness of supplied material, the allowed scope for the roughness of the finished products, andthe allowed scope for the specific elongation. The method can fully satisfy the customers in product mechanical property, surface quality, and the like, enhance the adaptability of the mill set to the products, lower the investment cost, exert the potential of the mill set and promote the use ratio of the device. The method of the invention is simple, clear and is suitable for on-line application.

The invention belongs to the technical field of the reactive voltage control over a power grid and more particularly relates to a reactive voltage verification method based on a limit operation mode. According to the method, firstly, through reading the data information of the power grid, four operation modes are formed, and then a limit operation mode is selected based on the calculation result of a comparison voltage. Secondly, a substation of the worst voltage exceeding condition is selected, and whether the two-side / three-side voltage of a substation transformer exceeds a voltage upper limit Umax or a voltage lower limit Umin or not is examined. Thirdly, a compensation factor epsilon is set and the amount of reactive power compensation QC / QL of the substation is calculated. Fourthly, whether the sum of the amount of reactive power compensation QC / QL and the actual input QCr / QLr of the reactive power compensation exceeds a pre-configured capacity QCmax / QLmax of the reactive power compensation or not is judged. Fifthly, the voltage is re-calculated based on the above adjustment. The above process is circularly conducted till no voltage exceeds the upper / lower limit. According to the invention, based on the calculation of limit operation modes, whether the configuration status of reactive power compensation equipment meets the requirements of the relevant provisions and the requirement of voltage adjustment or not can be effectively verified, so that the decision-making process of the power grid operation is guided. Meanwhile, the blank in reactive voltage checking method is filled, and the above method is more applicable to the actual operation status of the power grid. Therefore, the method is suitable for online application.

The invention discloses a peak value power estimation method for a power battery pack, solving the problem of not considering inconsistency between battery cells in the power battery pack, and havingthe advantages of significantly improving accuracy for estimating the peak value power of the power battery pack. The peak value power estimation method for a power battery pack includes the steps: updating a first-order RC equivalent circuit model parameter of a power battery in real time by means of an online identification method; estimating a power battery state of charge SoC on line, based ona state space expression of the first-order RC equivalent circuit model of the power battery; considering the state of charge SoC, and limitation of voltage, current and power of the power battery cells, calculating the peak value power of the power battery cells; and calculating the peak value power of the power battery pack according to the obtained peak value power of the power battery cells,wherein the peak value charge power is decided by the power battery cell with the highest voltage and the peak value discharge power is determined by the power battery cell with the lowest voltage.

The present invention relates to a method and device for detecting faults of principal component correlation degree sensors based on density clustering, wherein the method includes: judging the working condition data of the monitoring sensors by using the multi-working condition model of the unit, classifying the monitoring sensors, and obtaining the working conditions Condition cluster; Wherein, the working condition information of all monitoring sensors in the described working condition cluster constitutes matrix X; Utilize the working condition information of any two monitoring sensors in the described working condition cluster to form matrix K; Carry out principal component analysis to matrix K, obtain The main characteristic T of matrix K; Utilize the matrix X and main characteristic T after the normalization process to determine the degree of correlation of any two monitoring sensors; The degree of correlation is compared with a threshold, and the monitoring corresponding to the degree of correlation greater than the threshold The sensor performs fault detection.

The invention discloses an energy management strategy of a hybrid system based on reverse deep reinforcement learning. The strategy includes: using the optimization solution method to calculate the globally optimized SOC result as expert knowledge; creating a reward neural network; using reverse reinforcement learning to learn expert knowledge to obtain parameters of the reward neural network; creating an action neural network and evaluating the neural network; setting the vehicle SOC value before interaction; input the obtained SOC value before interaction into the reward neural network to obtain the reward value; input the obtained SOC value before interaction into the action neural network to obtain the mode allocation ratio; use the mode allocation ratio to interact with the environment, and obtain SOC value after interaction; input the SOC value before interaction, mode allocation ratio, reward value, and SOC value after interaction into the evaluation neural network to obtain the evaluation value; the agent calculates the gradient of each network and backpropagates to update the network parameters until the training is completed. The invention can learn the optimal reward function from expert knowledge, so that the effect of deep reinforcement learning is better.

The invention discloses a short-circuit and low-capacity fault diagnosis method, system and equipment of a series battery group that integrates balance information, and belongs to the field of circuit fault diagnosis, aiming at solving the problem that the short-circuit fault and low-capacity fault diagnosis process in the prior art cannot be considered in the circuit The effect of equalization cannot be applied to the flawed technical problem in battery systems equipped with equalization control. Multi-fault diagnosis of the battery pack in the balanced state performs fault detection according to whether the battery pack meets the consistency requirements. Fault detection realizes real-time detection and monitoring of short-circuit and low-capacity faults by analyzing the balance information under driving conditions and charging conditions. Distinguish and realize fault quantification, which solves the problem that the existing short circuit or aging-related fault diagnosis methods fail under equalization. Using the equalization process information to detect and distinguish short-circuit and low-capacity faulty cells in the battery pack, effectively reducing the calculation load of fault diagnosis and facilitating the application of the actual battery management system.

The invention relates to a plate shape closed-loop control method of cold-rolled strip steel based on influencing self learning of a matrix. An influence matrix empirical value table with different plate belt materials and different passes is established by determining two key influence factors; and in the process of the plate shape closed-loop control, actually-measured plate shape data is utilized to improve the quality of the influence matrix empirical value table continuously by a self-learning mode, so as to lead the influence matrix empirical value table to be closer to the actual regulating efficiency of a plate shape regulating mechanism. As the self learning process is carried out under the condition of the practical coupling action of influence factors of the regulating performances of all regulating mechanism, the method is more comprehensive than the factors considered by the intelligent method in a sense. In addition, in the process of closed loop control, influences on calculation of the matrix and self learning of the matrix are realized by adopting a simple mathematical algorithm, the calculating speed is high, and the real-time performance is good. Therefore, the method has the advantages that the reliability is high, the adaptability is strong, online application is suitable, the implementation is convenient and the like.

The invention provides a skip reentry aerial vehicle online trajectory planning guidance method based on a linear pseudo spectrum. The method comprises the following steps of initializing, performingstandard control on a flight segment, performing skip segment tilting counter-rotating determining, performing skip segment predictive trajectory integrating, performing precision determining, correcting skip segment rolling counter-rotating time correction, re-establishing parameter control, performing skip segment predictive trajectory integrating, performing precision determining, performing skip segment control parameter correction, and performing Kepler trajectory flight. According to the skip reentry aerial vehicle online trajectory planning guidance method, a roll angle modulus is updated through iteration and furthermore a terminal range deviation is corrected. When a segmentation condition is not decided by time, control compensation is introduced, thereby supplying a good initialcondition for terminal guidance. Furthermore a roll angle modulus value control rule and roll angle counter-rotating time are synchronously updated in one guidance period. Furthermore terminal transverse range and longitudinal range precision are simultaneously ensured. Through an analysis relation between a control amount correction value and a terminal offset, no iterative integration is required in updating the control amount at each time. Only 2ms are required in calculation at each time. The skip reentry aerial vehicle online trajectory planning guidance method is particularly suitable for online application.

The invention relates to a method for diagnosing sensor faults in a thermodynamic system of a thermal power plant, comprising: sampling sensor information data of the thermal system; determining the order p according to the sensor information sampling data; selecting an autoregressive model according to the order p; determining the autoregressive parameters in the model, and at the same time determine the expression of the autoregressive model; use the autoregressive model to predict the output value of the sensor at the next moment, and generate a residual by comparing the predicted value of the sensor with the actual value to obtain the residual sequence; The error sequence is analyzed, and when all the residual sequences show that there may be a certain type of fault, it is judged that the sensor has this type of fault. The technical solution can quickly and accurately realize sensor fault diagnosis and data reconstruction, and is especially suitable for online applications.

The invention discloses a method for detecting power grid frequency by using an optimal multiplier Newton algorithm. The method is characterized by being capable of simultaneously detecting a peak value of power grid voltage, a power grid frequency, a direct current component and an initial phase angle. The calculation is accurate when the frequency is perturbed greatly, the algorithm is simple, the convergence speed is high, the anti-jamming capability is strong and the method is suitable for online application.

The invention provides a short-circuit current limiting method based on power grid topological adjustment. The method includes the specific steps that 1, a potential strategy set is constructed; 2, candidate topological adjustment strategies are determined according to comprehensive effectiveness index values; 3, safety analysis is carried out on power grid short-circuit currents and the power grid operating state after the candidate topological adjustment strategies are executed; 4, transmission loss analysis is carried out on the power grid operating state, and a candidate topological adjustment strategy corresponding to the transmission loss minimum value is selected as a target topological adjustment strategy. Compared with the prior art, the short-circuit current limiting method based on the power grid topological adjustment has the advantages that the comprehensive effectiveness index values are adopted to objectively reflect limiting effects of the topological adjustment strategies on the whole network short-circuit currents, effective strategies can be quickly and accurately searched out of numerous potential strategies, safety assessment is conducted on the strategies, and operable topological adjustment strategies are determined.

The invention belongs to the technical field of reactive voltage control of a power grid, in particular relates to a layered and zoned reactive voltage analysis method based on a boundary condition. The layered and zoned reactive voltage analysis method comprises the following steps of making the power grid layered and zoned; determining a researched zone and a boundary thereof; inputting maximum daily load, minimum daily load, output of a power generator, the boundary condition, a reactive compensation configuration condition and a limit value of a regional power grid; carrying out power flow calculation, adjusting reactive compensation capacity, and adjusting the boundary of the regional power grid to the boundary condition; judging whether an out-of-limit situation exists in the voltage in a region or not; and judging whether a power factor of each main transformer conforms to the stipulation of the limit value or not. According to the layered and zoned reactive voltage analysis method, reactive voltage correction and reactive voltage optimization can be fulfilled, the reactive voltage analysis efficiency of a large-sized power grid can be substantially improved, focused analysis can be conducted with regard to a local power grid, and the layered and zoned reactive voltage analysis method is suitably used for situations with relatively complete data of the local power grid and insufficient data of a large power grid; and the layered and zoned reactive voltage analysis method is easy to understand and implement, is proximate to actual running of the power grid, and is suitably for on-line application.

The present invention discloses a multi-stage phishing website detection method and system, and combines means of both fast filtering and accurate filtering. Multiple stages of fast filtering are used to control the number of potential phishing websites to be in a relatively small range; furthermore, an accurate determination model is trained by analyzing statistical features of positive and negative samples in a small range. The method comprises the following steps: selecting a to-be-detected range of websites to perform fast filtering and excluding obvious non-phishing websites therefrom; and performing accurate determination on the remaining range of the websites after the fast filtering to determine whether said websites are phishing websites. The system comprises: a fast filtering module, configured to select a range of to-be-detected websites to perform fast filtering and exclude obvious non-phishing websites therefrom; and an accurate determination module, configured to perform accurate determination on the remaining range of the to-be-detected websites after the fast filtering.