4results about How to "Lower average wait time" patented technology

The invention discloses a double-elevator group control scheduling method and system based on scene self-adaption and machine learning, and belongs to the technical field of intelligent building control. In order to solve the problems that an existing double-elevator system is lagged in response in a non-peak period and low in peak period transport capacity matching degree, full-floor balanced response is achieved by monitoring the elevator state in real time, controlling double elevators to stop at the bottom layer in an idle mode and calculating a middle floor through a weighted gravity center method; in the peak mode, differentiated directional elevator parking strategies are executed according to building types, and the up-down tidal passenger flow of morning and evening peaks is accurately adapted. In addition, the system introduces a machine learning module, and dynamically updates a peak time window based on historical data by using a time sequence density detection algorithm. According to the method, the average waiting time is effectively shortened, the carrying efficiency in the peak period is improved, and the elevator system is endowed with the self-evolution ability adapting to the flow change of building personnel.

The application provides a simulation running time prediction method and system, a storage medium and a program product, and relates to the technical field of model training. The original running time is logarithmically transformed to compress the time data into a smooth numerical space that is easier for model learning. Data cleaning provides a clean and representative data basis for model training. Standardization processing avoids interference of different parameter dimension differences on model weight learning. The convolutional neural network model has local perception ability and can automatically capture and learn deep patterns of strongly coupled parameter combinations from the standardized feature vector. Finally, the introduction of the Huber loss function ensures that the model can accurately fit normal samples and resist the interference of residual noise during the training process, and finally obtains a highly robust and generalizable prediction model.

The application discloses a kind of based on multi-objective optimization's charging cabinet resource scheduling system, system includes data perception layer, intelligent decision-making layer and execution control layer, the intelligent decision-making layer integrates demand prediction unit, multi-objective optimization unit and reinforcement learning dynamic adjustment unit, form " perception-prediction-static optimization-dynamic adjustment-execution-feedback " intelligent closed loop, wherein, demand prediction unit adopts space-time convolutional neural network, realizes accurate space-time prediction to charging demand;Multi-objective optimization unit constructs and solves multi-objective model, cooperatively optimizes charging cabinet utilization, user waiting time and power grid load fluctuation, generates Pareto optimal scheduling scheme set;Through deep reinforcement learning intelligent agent, online fine-tuning is carried out to basic scheme, to deal with emergency situation.The application realizes the forward-looking planning of charging resource scheduling, multi-objective balance and dynamic robustness adaptation, can significantly improve the overall operation efficiency of charging network, user experience and power grid compatibility.

This invention provides a single-channel video carousel method, apparatus, electronic device, and storage medium. The single-channel video carousel method includes: determining that the channel bandwidth is k times the video playback bitrate, where k ≥ 1, based on the available video carousel bandwidth and the playback bitrate of the target video; using a video segmentation algorithm to simultaneously determine the number of video segments N, where N ≥ k, and the position of each segment in the channel, based on k; and dividing the video into N equal-length video segments, S... i Let S be the i-th video segment, 1≤i≤N, and let S be the total number of video segments. i A complete video is formed by concatenating the segments in ascending order of their sequence numbers. Based on a video segmentation scheme, corresponding segments are deployed at the corresponding positions on the channel for cyclical transmission, achieving video data rotation. The video data reception method includes: assuming the client user requests video playback at time t... a Based on the client receiving algorithm according to t a Determine the reception time and video playback time for each segment.