40 results about "Incremental learning algorithm" patented technology

The invention relates to a delivery vehicle scheduling method with time limitation, and belongs to the technical field of vehicle routing in intelligent optimized scheduling. A scheduling model and an optimization object in the delivery process with time limitation are determined, the optimization object is obtained by solving the scheduling model on the basis of an improved population incremental learning algorithm, the optimization object of the scheduling model is the minimal total transport cost, and the limitation condition is established according to time limitation of the delivery vehicle, load-carrying restriction, position coordinates of a client point and the like. Global searching is guided by a possibility matrix observation model, individual information in a historical group is optimized in the improved population incremental learning algorithm, global searching is effectively guided, and the efficiency of global searching is higher; global searching aimed at problem property is constructed based on limitation of the problem, a high-quality solution area obtained via global searching is rapidly and effectively searched, the performance of the algorithm is improved, and the optimal solution is obtained.

The invention discloses a relevance vector regression incremental learning algorithm and system based on sample characteristics. The method comprises the following steps of: S1, obtaining an initial sample set, and initializing parameters; S2, training the sample set to obtain an RVM prediction model; S3, calculating a sample label, a local density factor and an error factor of each sample; S4, predicting a future sample to be input according to the RVM prediction model; S5, calculating sample characteristic vectors, arranging the sample characteristic vectors in a descending order, performing circulation, if the time of non-relevance vectors is beyond a set threshold value, deleting the sample from the sample set, and exiting the circulation; and S6, judging whether a new input sample exists or not, if so, adding a new sample to form a new sample set, turning to the step S2, and if not, outputting the predicted future sample. By means of the relevance vector regression incremental learning algorithm and system disclosed by the invention, a sample including effective information can be reserved; an invalid sample can be deleted; the prediction precision is relatively high; the time complexity is relatively low; and thus, the relevance vector regression incremental learning algorithm and system can be widely applied in real-time data processing and prediction.

The invention discloses a virtualized network function placement method based on a population-based incremental learning algorithm. The population-based incremental learning algorithm is applied to apreset network topology, minimization of service delay is taken as an optimization target, and a virtualized network function placement scheme satisfying a practical deployment demand is calculated. The method is mainly characterized by that a two-stage coding mode is employed, reparation operation is carried out on illegal individuals, a Floyd algorithm is employed when individual fitness is calculated, a probability vector is guided to be updated through storage of a global elite set, the probability vector is changed through mutation operation, and the like. According to the method, the PBIL (Population-Based Incremental Learning) algorithm is applied to solution of the VNF-P (Virtualized Network Functions Placement) algorithm; simulation experiment and data analysis show that comparedwith a genetic algorithm, the method provided by the invention has the remarkable advantages of algorithm performance and efficiency; a solution enabling service delay to be relatively small can be obtained; and the feasibility and high efficiency of the method are proved.

The invention provides a method for controlling standby restart of a centrifugal-type water cooling unit. The method includes step S1, acquiring running parameters before standby, and comparing the running parameters with preset parameters to determine a standby reason; step S2, when the standby reason is water temperature standby, setting restart target parameters according to an incremental learning algorithm; step S3, controlling restart of the unit according to the restart target parameters. According to the method, when the standby reason is water temperature standby, the restart target parameters are set according to the incremental learning algorithm, and restart of the unit is controlled according to the restart target parameters, so that running load of the unit is enabled to be matched with working conditions, the unit is enabled to run under small-load working conditions, potential safety hazards caused by frequent start and stop of the unit are lowered, and the unit is enabled to be more energy saving. In addition, high-low pressure difference of the unit can be lowered, so that the unit is enabled to be away from a surge area, and potential surge possibility of the unit is lowered.

In view of how to analyze the operating state of a transformer and perform fault diagnosis based on a vibration method, the embodiment of the invention discloses an online abnormality detection methodfor transformer vibration. The method comprises: obtaining new data samples about a vibration signal of the transformer operating state within a preset time period; extracting the characteristic parameters of the new data samples based on wavelet packet analysis; completing the learning of the new data samples based on a fast convex hull algorithm, training and updating a single-class abnormalitydetector model; and calling the updated single-class abnormality detector model to perform online abnormality detection of the current transformer vibration. According to the online abnormality detection method for transformer vibration, since the incremental learning algorithm is used to learn only the new data samples within the preset time period, the new data can be processed in time and effectively, the new training samples can be learned online in real time, and rapid upgrade of detection models can be achieved. At the same time, time and space requirements of the model update can be reduced.

The invention discloses a face recognition method based on improved incremental non-negative matrix decomposition, which adopts an incremental non-negative matrix solution algorithm based on Fisher discriminant analysis. The algorithm uses the prior information of the initial sample training, assigns the column vector of the new coefficient to the mean vector of the corresponding category through the index matrix, and improves the convergence speed of the algorithm; in addition, the INMF algorithm is improved to a batch incremental learning algorithm , use the index matrix for initialization assignment, and impose the constraints of maximum inter-class scatter and minimum in-degree on the objective function, so as to obtain the best subspace projection. This solution has a higher recognition rate and a faster recognition speed, and is suitable for occasions such as face recognition.

The invention discloses a method for constructing a knowledge base based on incremental learning and generating a partial sequence structure graph. It adopts the idea of ​​incremental learning based on the covering principle to guide the covering operation of attributes and object sets, and completes special functions such as attribute libraries through special covering relations. The construction of the set further completes the generation of the formal background, and completes the deletion of redundant patterns; it is constructed on the basis of the formal background, and its computer generation algorithm obtains the precise coordinate positioning of each concept node by calculating the hierarchical pattern matrix. Layers, nodes, connections and other elements, as well as the introduction of incremental learning algorithms, complete the dynamic construction of partial order structure graphs. Compared with the prior art, the present invention combines the incremental learning based on the coverage principle and the construction algorithm of the partial order structure graph, which can realize the dynamic construction of the knowledge concept base and complete the automatic generation of the hierarchical pattern matrix; realize the partial order structure Incremental construction of graphs.

The invention discloses a virtualized network function placement method based on a population-based incremental learning algorithm. The population-based incremental learning algorithm is applied to apreset network topology, minimization of service delay is taken as an optimization target, and a virtualized network function placement scheme satisfying a practical deployment demand is calculated. The method is mainly characterized by that a two-stage coding mode is employed, reparation operation is carried out on illegal individuals, a Floyd algorithm is employed when individual fitness is calculated, a probability vector is guided to be updated through storage of a global elite set, the probability vector is changed through mutation operation, and the like. According to the method, the PBIL (Population-Based Incremental Learning) algorithm is applied to solution of the VNF-P (Virtualized Network Functions Placement) algorithm; simulation experiment and data analysis show that comparedwith a genetic algorithm, the method provided by the invention has the remarkable advantages of algorithm performance and efficiency; a solution enabling service delay to be relatively small can be obtained; and the feasibility and high efficiency of the method are proved.