9results about How to "Reduced cold start time" patented technology

The present application relates to a kind of preparation methods of essence.The problems that traditional process parameter adjustment relies on artificial experience and is difficult to realize cross batch self-adaption, multi-modal feature extraction, parameter-function-stage three-dimensional label matrix generation, graph neural network behavior atlas construction and expert strategy distribution constraint dual-channel strategy learning method are proposed.By process parameter and derived feature multidimensional fusion, time series causal strength analysis and action instruction simulation verification, optimization process parameter combination with explainable function synergistic relationship can be output.The technical scheme effectively realizes the automatic optimization of process control decision and the rapid self-adaption of model under new raw material batch, improves product consistency, quality stability and control energy efficiency.

The invention discloses a container mirror image preloading method and device based on demand prediction, equipment, a medium and a product, and the method comprises the steps: collecting a current node state, a task execution state and mirror image information in real time, and carrying out the data preprocessing; inputting the preprocessed data into a preset hybrid prediction model, and outputting the demand probability of each node for each container mirror image in a future time window; the hybrid prediction model comprises a time sequence feature extraction network used for learning a time sequence rule and an attention coding network used for learning global relation features; determining a to-be-loaded container mirror image and a target node according to the demand probability and the resource constraint of the node; and preloading the to-be-loaded container mirror image to the target node. According to the method, the demand probability of the container mirror image on the node is predicted, only the necessary container mirror image is preloaded, the cold start time of the container is shortened, the resource utilization rate of the node is optimized, the consumption of the storage space and the network bandwidth is reduced, and the operation cost is effectively reduced.

The invention provides a panoramic image display method and device, a vehicle and a storage medium. The method comprises the steps that a state machine thread and an image processing thread of a panoramic image process are initialized; if the state of the panoramic image application is an activated state, initializing a panoramic image display window and creating a display carrier through the state machine thread, and initializing a shader through the image processing thread; and processing the display carrier through the image processing thread to create a rendering surface, processing image data, rendering the image data to the rendering surface, and displaying a panoramic image. In the embodiment of the invention, different tasks in the panoramic image display process are distributed to independent threads, and the threads process the tasks in a parallel manner, so that the operation efficiency and response speed of the panoramic image application are improved, the cold start time of the panoramic image application is shortened, and the user experience is improved.

The utility model relates to a megawatt anion exchange membrane water electrolysis hydrogen production integrated system which comprises an electrolytic bath subsystem comprising a plurality of large standard AEM electrolytic baths and is used for water electrolysis hydrogen production. The electrolyte circulation subsystem is used for providing electrolyte and maintaining circulation; the oxygen treatment subsystem is used for drying and dehydrogenating oxygen; the drying and purifying subsystem is used for drying, deoxidizing and purifying the hydrogen; the purging subsystem is used for purging the residual gas; the power supply and distribution subsystem is used for supplying power; the heat dissipation subsystem is used for cooling and heat dissipation; and the centralized control subsystem is used for controlling the whole integrated system. According to the utility model, the centralized control subsystem is adopted to realize remote control and real-time monitoring on the whole integrated system, so that the whole integrated system is in an optimal operation state, and the efficiency and the dynamic response rate are effectively improved; according to the utility model, the purging system is adopted to purge residual gas, so that the safety of the system is improved; and the low-concentration alkaline liquid is adopted, and hydrogen is removed from oxygen, so that the environmental protection property is effectively improved.

This invention discloses a container image preloading method, apparatus, device, medium, and product based on demand prediction. The method includes: real-time acquisition of current node status, task execution status, and image information, and performing data preprocessing; inputting the preprocessed data into a preset hybrid prediction model to output the demand probability of each node for each container image within a future time window; the hybrid prediction model includes a time-series feature extraction network for learning time-series patterns and an attention encoding network for learning global relationship features; determining the container image to be loaded and the target node based on the demand probability and the node's resource constraints; and preloading the container image to be loaded onto the target node. This invention, by predicting the demand probability of container images on a node, preloads only the necessary container images, shortening container cold start time while optimizing node resource utilization, reducing storage space and network bandwidth consumption, thereby effectively reducing operating costs.

This invention discloses a starting control method, system, equipment, aircraft, and medium for a propeller-driven aircraft piston engine, relating to the field of aero-engine control technology. The method includes: performing a power-on self-test before starting, checking the minimum propeller blade angle, throttle opening, and fuel pressure; after the self-test, the flight control computer automatically and adaptively controls the starting throttle opening and target starting speed based on altitude and atmospheric static temperature; during starting, status monitoring is performed, towing the engine to a state capable of ignition and fuel injection control; after successful starting, the lubricating oil temperature is determined to automatically set the corresponding throttle opening and warm-up speed; finally, a dual ignition system check and pre-flight condition checks are performed. This invention achieves fully closed-loop adaptive control, effectively overcoming starting difficulties in complex environments such as high altitudes and frigid conditions, eliminating the hidden dangers of starting with defects, and significantly improving the timeliness of the aircraft's maneuverability and operational safety.

This invention proposes a causally explainable anomaly analysis and root cause tracing method, apparatus, and readable storage medium, aiming to solve the problems of poor anomaly identification stability, root cause localization bias, and insufficient diagnostic interpretability caused by non-stationary data, small sample sizes, and dynamic evolution of process logic in rapid batch production changeover scenarios. This invention collects and preprocesses multi-source heterogeneous data, monitors production indicators based on adaptive thresholds and extracts anomaly window feature data, migrates the initial causal structure from historical batches based on multi-dimensional similarity of product model, process route, and anomaly pattern, and performs dynamic attribution updates combined with online data. Based on a causal graph, it applies counterfactual perturbations to candidate factors to calculate causal contribution and searches backward along the causal chain to pinpoint the root cause, mapping the inference path to physical semantics to generate a diagnostic report. This invention is applicable to anomaly diagnosis in multi-product, rapid-change flexible production lines.

The utility model relates to the technical field of vehicle chassis, and provides an exhaust catalyst and a vehicle. The exhaust catalytic converter is arranged in an exhaust system connected with an engine and comprises a catalytic converter body with a catalytic unit and an energy storage assembly arranged on the catalytic converter body. Wherein a phase change material and a heat exchange pipeline are arranged in the energy storage assembly, the heat exchange pipeline is connected with a gas outlet pipeline of the exhaust catalyst, meanwhile, when the engine works, the heat exchange pipeline can introduce gas in the gas outlet pipeline to heat the phase change material, and when the engine stops, the phase change material can release heat to heat the catalytic unit. By means of the heat storage and heat release functions of the energy storage assembly, the catalytic unit can be maintained at a certain temperature when the engine stops, the catalytic unit can reach the ignition temperature as soon as possible when the engine is in cold start, the cold start time of the engine can be shortened, emission exceeding during cold start can be reduced or even avoided, and the service life of the engine is prolonged. And the use quality of the vehicle can be improved.

The invention relates to the technical field of station reactive power response, in particular to an Internet of Things edge device scheduling system and method based on AI reasoning hardware resources, and the system comprises an edge system module, a cloud storage module, an ML model data module and a reasoning data module. A computer cluster, an artificial intelligence reasoning service unit and a local storage unit are integrated in the edge system module; the edge system module, the cloud storage module, the ML model data module and the reasoning data module communicate with one another through the network service unit; wherein the edge system module is used for receiving Internet of Things equipment data at an edge side and executing an AI reasoning task, and providing a reasoning result for an upstream application with relatively low delay; and the cloud storage module is used for storing model backup, historical reasoning data and logs and providing cloud reasoning rollback capability. According to the invention, through intelligent sensing and scheduling of heterogeneous edge hardware, efficient deployment of the ML model in the resource-constrained IoT environment is realized.