29 results about "Thermal aware" patented technology

The invention discloses an electromagnetic heating system, and a method and a device for detecting the thermistor of the electromagnetic heating system. The method comprises the following steps of: recording the temperature value obtained by the thermistor as a first temperature value when the electromagnetic heating system is powered on; recording the temperature value obtained by the thermistoras a second temperature value when the electromagnetic heating system is in a standby state; obtaining the maximum temperature value and the minimum temperature value of the temperature value obtainedby the thermistor in a continuously preset second preset time every other first preset time during the heating process of the electromagnetic heating system; and judging whether the thermistor is ina failure state or not according to the first temperature value and / or the second temperature value, the maximum temperature value, and the minimum temperature value. The method for detecting the thermistor of the electromagnetic heating system judges whether the thermistor fails or not according to the static temperature values at the time of power-on and standby and the dynamic temperature valuein the heating process, so that the accuracy and the reliability of the judgment are greatly improved, and the safety of the system is improved.

The application discloses a method for unmanned aerial vehicle assisted thermal sensing MEC network task scheduling and resource allocation. In the case of considering CPU thermal constraints, the cruising time of the unmanned aerial vehicle is minimized by jointly optimizing user admission policy, task scheduling, unmanned aerial vehicle hovering trajectory, flight time, and computing and communication resource allocation. The optimization problem is further decomposed into three sub-problems and iteratively processed in order. Simulation results show that the method has good performance improvement effect compared with other algorithms.

The invention provides a thermal-sensing hybrid task workflow collaborative scheduling method and system in a software defined satellite environment. The method comprises the following steps: acquiring a user request containing a plurality of sub-tasks, and establishing a dependency relationship among the sub-tasks based on an execution time sequence of each sub-task; allocating available time windows for the hardware sub-tasks in the user request, and organizing the software sub-tasks depending on the same hardware sub-task into an ordered batch topology; and obtaining performance parameters of the system when the hardware sub-tasks execute user requests in different time windows and / or the software sub-tasks are in different batch topologies, and obtaining an optimal sub-task scheduling strategy based on the performance parameters. According to the method, overheating and overload risks are actively avoided in the planning stage; and meanwhile, a task arrangement structure is collaboratively optimized, so that the traditional contradiction between high throughput and thermal safety is broken, and the utilization efficiency of satellite resources under complex constraints and the overall efficiency of the system are remarkably improved.

The invention discloses an infrared thermal imaging diagnosis system and method based on a converter station intelligent gateway, and the system comprises an infrared thermal imaging collection unit which is used for collecting the image data of a core thermal load region through arranging a visual and infrared thermal sensing terminal in the core thermal load region of to-be-monitored equipment, sending the data to the edge computing intelligent gateway; the edge computing intelligent gateway is deployed in a core thermal load area, and the edge computing intelligent gateway is used for preprocessing the image data and performing preliminary judgment on the preprocessed image data to obtain a preliminary thermal analysis result of the image data; generating alarm information based on the preliminary thermal analysis result, and sending the alarm information to an operation and maintenance management platform; the operation and maintenance management platform is used for fusing the image data in the alarm information with the multi-source data to generate fused data; and the operation and maintenance management platform is used for analyzing the fusion data through a deep learning image recognition model and outputting a diagnosis result of the to-be-monitored equipment.

The invention discloses a silicon bridge interconnection core particle thermal sensing layout method based on reinforcement learning. The method comprises the following steps: S1, receiving a configuration file containing a core particle set, a size, an identifier and a silicon bridge connection relation network; s2, constructing a reinforcement learning environment, setting the size of a substrate and the size of a grid, and generating a fixed placement sequence of core particles based on a connection relationship; s3, determining legal candidate positions in sequence, sampling a preset area of the substrate in an empty layout, taking an intersection of candidate positions of adjacent core particles in a non-empty layout, and filtering illegal positions; s4, the intelligent agent selects an action containing a placement coordinate and a rotation state from the legal action set for execution; s5, calculating a dual-stage reward, guiding a legal layout, and optimizing thermal performance, area and bus length; s6, neural network parameters are updated based on a reinforcement learning algorithm; and S7, circulating the steps S3-S6 until the termination condition is met, and outputting the optimal layout meeting the core constraint. According to the method, the exploration space is effectively reduced, the solving efficiency is improved, and the cooperation of the silicon bridge interconnection reliability and multi-objective optimization is realized.

The invention discloses a GPU (Graphic Processing Unit) chip temperature sensing thread bundle management method, relates to the technical field of GPU chip application, and solves the problems of low efficiency of a thermal management strategy and the like caused by the fact that a thermal prediction method in the prior art cannot cover hot spots of a whole chip, is high in modeling complexity and poor in real-time performance. The thermal prediction model is used for predicting the hot spot temperature of the chip in the running process of the graphics processor; the TAWS scheduling module is constructed based on a two-stage Warp scheduling strategy and is used for comparing the hot spot temperature with a preset maximum temperature threshold value and selectively executing a normal scheduling mode or a temperature control scheduling mode; the temperature of the chip is regulated and controlled; and when the duration of the high-temperature state exceeds the preset maximum allowable duration, the TAWS scheduling module triggers a scheduling freezing mechanism to realize safety control of the chip temperature. According to the method, cross-architecture high-precision generalization, a low-overhead instruction-level thermal sensing scheduling mechanism, refined analysis of instruction-level thermal characteristics and the like are realized.

The invention discloses a thermal sensing core particle layout method for silicon bridge interconnection, and relates to the technical field of packaging. According to the method, a multi-core particle layout optimization problem oriented to silicon bridge interconnection is formalized into a normalized mathematical programming model. Based on an existing mature mathematical solution method, a legal layout scheme meeting the silicon bridge interconnection constraint is rapidly obtained. Meanwhile, solution constraints are eliminated, differentiated legal layout solution sets can be generated through the constraints, then systematic thermal characteristic evaluation is carried out on different layout schemes, and finally a silicon bridge interconnection-oriented multi-core-particle layout scheme considering interconnection constraints and thermal performance optimization is obtained.

The invention relates to a point cloud thermal prediction method for a large-scale unstructured 2.5 D integrated circuit. The network comprises three core modules: an adaptive multi-path coupling diffusion module (AMD), a wavelet-based fine-grained recovery module (WFR) and a thermal sensing material expert hybrid adapter (TA-MoME). The AMD module adaptively learns interaction of a thermal diffusion path through an attention mechanism based on serialized gating; the WFR module recovers fine-grained thermal gradient features by using a high-frequency wavelet domain enhancement technology; and the TA-MoME adapter dynamically routes the material specific expert to adapt to the heat distribution difference between the heterogeneous materials. Particularly, the framework also introduces a continuous learning mechanism with efficient parameters to ensure that when the model adapts to a new geometric scale, catastrophic forgetting can be prevented, and cross-scale adaptation and mobility can be realized. Experimental results show that in 80K-scale point cloud prediction, compared with an existing FSA-Heat method, the average precision index is remarkably improved (errors are reduced by about 78% or above), compared with commercial COMSOL, acceleration of 147 times is achieved, and the generalization ability for larger-scale nodes (0.4 M) and zero samples without geometrical shapes is shown.

Hot-aware interrupt routing in an interrupt controller in a processor-based system and related methods are disclosed. The processor-based system includes one or more interrupt controllers each configured to prioritize interrupts received from components in the processor-based system and direct each received interrupt to a CPU core in a designated processor of the interrupt controller to execute an interrupt service routine (ISR) to process the interrupt. To avoid or reduce the likelihood that the interrupt controller directs an interrupt to a CPU core and / or CPU core cluster that can exceed its thermal limit by accepting and handling the interrupt, the interrupt controller is configured to be aware of the temperature of the CPU cores and / or CPU core clusters in its designated processor. The interrupt controller is configured to selectively route a received interrupt based on the temperature of the CPU core and / or its CPU core cluster that is determined to be eligible to receive and handle the interrupt.

The invention provides a non-contact thermal sensing system and method, and belongs to the technical field of robot sensing and flexible electronics. In order to solve the problems that physical contact is needed for temperature sensing of an existing robot, response is slow, and a heat source cannot be pre-positioned, the system comprises a flexible temperature sensor array composed of sensor units containing two-dimensional material film temperature-sensitive layers, and the flexible temperature sensor array has millisecond-level thermal response time; a non-thermal sensor; and a processor. The processor is configured to: acquire temperature data of the array and signals of the non-thermal sensors; judging whether a heat source approaches without physical contact based on the two signals; in response to the determination, a spatial location of the heat source is calculated based on a temperature distribution formed over the array. According to the invention, by fusing the multi-mode signals, rapid and non-contact pre-sensing and spatial positioning of the heat source are realized, and the operation safety of the robot is improved.

The invention discloses a global sensing type intelligent temperature control cable and a preparation method thereof, and the cable comprises a conductor, an inner insulating layer, an intelligent sensing response layer, an outer insulating layer, a fireproof layer, an electromagnetic shielding layer and a sheath layer, and the intelligent sensing response layer comprises a polyimide composite belt, a sensing micropipe and a liquid cooling micropipe. Perfluoropolyether cooling liquid and gallium-indium-tin liquid metal micro-droplets are filled in the sensing micro-tubes, high-thermal-conductivity nano-fluid is filled in the liquid-cooling micro-tubes, and the sensing micro-tubes and the liquid-cooling micro-tubes form a sensing channel and a temperature control channel through parallel design respectively; inlets and outlets of the sensing channel and the temperature control channel are arranged at the end part of the cable and are integrated with a double-cavity micro piezoelectric pump and a double-cavity micro radiator to form a sensing array and a temperature control array, and an embedded edge controller electrically connected with a sensing micro tube is integrated in the double-cavity micro piezoelectric pump. According to the invention, transmission, heat dissipation and sensing are integrated, and accurate positioning of various faults such as extrusion, cutting, overheating and the like is realized.

The invention discloses a heat sensing data center energy consumption and service life optimization method and system based on the health degree and a storage medium thereof, and belongs to the technical field of data center energy saving and equipment service life management. A sustainable updating server health degree scoring model is constructed, a temperature fluctuation memory window is introduced to describe a thermal fatigue effect, IT energy consumption, cooling energy consumption and life loss cost converted from health degree reduction are uniformly incorporated into a multi-time-period joint optimization model, and a two-stage time scale scheduling strategy is adopted to optimize the health degree of a server. On the premise that service requirements and temperature safety are guaranteed, the server operation environment and the load state are guided to evolve towards the direction of low energy consumption, low temperature impact and low failure rate, and collaborative optimization of the total energy consumption of the data center and the service life of the server is achieved. On the premise of meeting business requirements and temperature safety, comprehensive energy consumption is reduced, health degree attenuation is slowed down, the service life of the server is prolonged, and long-term operation stability is improved.

The application discloses a server BMC cooperative optimization method and system based on heat sensing and energy consumption prediction, relates to the technical field of server hardware management and data center energy saving, and discloses a server BMC cooperative optimization method and system based on heat sensing and energy consumption prediction, which comprises the following steps: a heat-energy coupling model is established through multidimensional data acquisition and fusion analysis; precise sensing is realized by adopting three-dimensional heat distribution calculation and LSTM prediction; heat dissipation and power supply parameters are dynamically adjusted in combination with a multi-objective optimization algorithm; a closed-loop feedback mechanism is formed; the energy efficiency ratio of the server can be improved; in addition, dynamic cooperative optimization of heat dissipation control and energy consumption management, improvement of server heat dissipation efficiency and prolongation of the service life of hardware can be realized.

The application discloses a kind of laser cutting machine rail searching path planning optimization method and system, it is related to laser cutting machine path optimization technical field, specifically includes the following steps: when the prediction result is that there is heat distribution overlap, all path groups with this condition are screened out, and it is calibrated as to be adjusted cutting path group;Real-time acquisition of the thermal sensitive characteristic information of each group of to-be-adjusted cutting path group, and after acquisition, analysis is carried out, the heat interference intensity generated by each group of to-be-adjusted cutting path group under current order is evaluated, and each group of to-be-adjusted cutting path group is divided into low heat interference path group, medium heat interference path group and high heat interference path group according to the evaluation result;According to the division result, corresponding cutting path order optimization measures are respectively executed.The application solves the problem that path heat overlap cannot be dynamically avoided, realizes intelligent optimization of cutting order based on heat sensing and accurate control of heat interference.

This invention provides a temperature derating method in thermally sensed electromigration (EM) analysis, which may include identifying a heat-sensitive structure, a first heat-generating structure, and a first heat-dissipating structure. The temperature derating method includes adjusting the evaluation temperature of the heat-sensitive structure by adjusting the operating temperature of the first heat-generating structure, and calculating the temperature increase of the heat-sensitive structure.

A processor-implemented method for thermal aware performance / power management for concurrent operation of large models includes monitoring a first temperature of one or more of the plurality of processors. Each of the one or more processors executes a machine learning model workload. The processor-implemented method also includes determining if the first temperature of the one or more of the plurality of processors is within one of a set of temperature zones. Power to the one or more of the plurality of processors is controlled based on the one of the set of temperature zones in which the first temperature of the one or more of the plurality of processors lies.

A machine-learning based, rapid, physics-aware thermal simulator, drawing inspiration from the Fourier's law and the Fourier-Biot equation, the first and second derivatives of the temperature map, is provided. The learning objective evolves from merely translating images to approximating natural phenomena such as the thermal gradient and thermal Laplacian. By adding an additional encoder during training and substituting the image-based loss with the thermal-aware loss, the proposed model achieves lower prediction error, higher data efficiency, and more physically accurate behavior.

The invention relates to the technical field of semiconductor thermal management, and discloses an edge computing task scheduling system of a diamond-based semiconductor device, which comprises an instruction preprocessing module, a thermal sensing control module and a scheduling execution module. The instruction preprocessing module reorganizes the instruction sequence into micro-slices and marks the micro-slices as a high-power burst type or a low-power gap-filling type. The thermal sensing control module calculates a predicted junction temperature based on the real-time voltage and current using a transient thermal impedance model. The scheduling execution module regulates and controls hardware according to a comparison result of the junction temperature and a safety threshold; when the junction temperature is safe, burst slicing is executed; when the junction temperature reaches the limit, the clock frequency and the voltage are kept unchanged, and a gap-filling slice is inserted and executed. According to the invention, the rapid thermal relaxation characteristic of the diamond substrate is utilized, active cooling of the device is realized through instruction flow gap filling on the premise of no frequency reduction and no voltage regulation, the switching overhead of the traditional frequency and voltage regulation technology is avoided, and the energy efficiency ratio and the thermal stability of the system are improved.

The application discloses a kind of AGC instruction dynamic distribution methods of fusing boiler heat storage state, it is related to thermal power generation control technical field, the present application includes the following steps: using pre-training parameters, the flue gas temperature of furnace outlet, water-cooled wall outlet working medium temperature, steam-water separator pressure and main steam and feedwater flow difference are mapped into nonlinear feature vector that characterizes heat storage space form feature and antagonistic latent heat state;Receive AGC instruction and actual power, extract heat storage perception from feature vector, calculate the equivalent load deviation of compensation heat storage, generate virtual guide rail;The time gradient of feature vector is calculated to determine the potential energy direction factor, generate instruction: when indicating violent confrontation, boiler proportional response deviation, steam turbine inhibition and feedforward;Indicate stable migration, boiler integral response, steam turbine fast tracking rail;Predict future main steam pressure, compare with high and low limit value and clamp gate and combustion rate instruction;Output clamped instruction, store parameters and vector into history window.

The invention provides an AI-based quick charger and an intelligent temperature control system and method thereof. Belongs to the technical field of intelligent power management. The method comprises the following steps: carrying out multi-mode thermal sensing area division on the quick charger to generate charger thermal sensing area data; a multi-dimensional thermal sensing network is constructed, temperature distribution data and battery impedance change data of key parts are collected in real time, and original multi-modal thermal data are generated; according to the original multi-mode thermal data, combining a pre-established device aging model and an environment parameter database; performing dynamic weight distribution, generating comprehensive thermal state evaluation data, adaptively adjusting a charging power threshold value, and synchronously triggering heat dissipation strategy selection to obtain a power adjustment instruction and a heat dissipation execution instruction; through the multi-mode thermal sensing network, the thermal field distribution and hot-spot evolution of the charger can be comprehensively and accurately captured, the accuracy of temperature monitoring is improved, and a reliable basis is provided for intelligent temperature control.

In some aspects, timing metrics for timing paths of a circuit design are accessed. These timing metrics were evaluated at a first temperature. The circuit design includes multiple components. Location-dependent temperature information for the components and a thermal derate schedule are also accessed. The thermal derate schedule specifies adjustments to timing metrics as a function of temperature. A processing device adjusts the timing metrics based on the thermal derate schedule and based on differences between the first temperature and the location-dependent temperature information for the respective components of the circuit design.

A semiconductor device includes a first substrate. The semiconductor device includes a plurality of metallization layers formed over the first substrate. The semiconductor device includes a plurality of via structures formed over the plurality of metallization layers. The semiconductor device includes a second substrate attached to the first substrate through the plurality of via structures. The semiconductor device includes a first conductive line disposed in a first one of the plurality of metallization layers. The first conductive line, extending along a first lateral direction, is connected to at least a first one of the plurality of via structures that is in electrical contact with a first through via structure of the second substrate, and to at least a second one of the plurality of via structures that is laterally offset from the first through via structure.

A heat sensor (1) comprises: a substrate (40); a heat detection element (30) that detects heat; and a housing (10) that accommodates the substrate (40). The substrate (40) includes a main body part (41) and an extension part (42) which extends from a peripheral edge (41a) of the main body part (41) to the outside of the main body part (41), and on which the heat detection element (30) is mounted. The extension part (42) has a mounting region (42a) on which the heat detection element (30) is mounted, and a pair of connection regions (42b) connected to the main body part (41). A hole (43) penetrating the substrate (40) is formed by the peripheral edge (41a) of the main body part (41) and the extension part (42). A first distance (L1) between the pair of connection regions (42b) is longer than a second distance (L2) from the center (O1) of the mounting region (42a) to the peripheral edge (41a) of the main body part (41).

The invention discloses a chip power consumption optimization system based on a generative adversarial network, and the system comprises the following modules: a data preprocessing module which is used for obtaining and processing chip design data; the thermal sensing logic diagram generation module is used for generating a thermal sensing logic diagram based on the thermal sensing memory migration network; the graph deformation optimization module is used for calculating an edge weight of a logic connection edge and executing deformation iteration processing through an attractive repulsion driving deformation mechanism; the critical path protection module is used for generating a limited optimization circuit diagram based on a synaptic logic retention mechanism; the multi-target discrimination module is used for outputting a discrimination feedback result based on a multi-target discriminator; the constraint adjustment module is used for generating updated EDA design constraint data; and the rear-end optimization implementation module is used for importing the limited optimization circuit diagram and the updated constraint data into a rear-end physical design tool for wiring. According to the method, the thermal sensing memory migration network and deformation iterative optimization are fused, and the chip power consumption optimization layout is realized.

The invention relates to the technical field of industrial energy conservation and waste heat recovery, in particular to an isooctane production waste heat cascade recycling device and a control method. The waste heat sensing and collecting layer is used for monitoring multi-strand waste heat source parameters in the isooctane production process in real time; the stepped conversion refining layer is connected with the waste heat sensing and collecting layer and used for carrying out stepped conversion on the waste heat according to the grade of the waste heat; through multi-stage cascade connection, each stage of working medium pair works in the most efficient temperature interval, the utilization rate of wide-range waste heat is maximized, and the loss is remarkably reduced; the system pressure can be accurately predicted and controlled, the problem of inter-stage pressure balance of a multi-stage system is solved, and stable and efficient operation of the system under various working conditions is ensured.

The invention discloses a core particle thermal sensing layout planning method based on rapid thermal analysis and reinforcement learning, and belongs to the field of electronic design automation. The method comprises the following steps: expressing layout planning as a Markov decision process by adopting reinforcement learning so as to jointly optimize a bus length and a peak temperature; in order to solve the problem that traditional thermal evaluation is time-consuming, a rapid thermal model of physical information is integrated, and the model calculates peak temperature at a high speed by pre-representing a self-thermal impedance and mutual thermal impedance lookup table and combining an interpolation method; in order to enhance the generalization ability, a multi-modal fusion architecture is adopted, the architecture uses a convolutional neural network in parallel to process the visual modality of the layout, and a graph neural network is used to process the topological graph modality of the core particles; in order to solve the sparse reward problem in reinforcement learning training, a customized dense reward strategy is adopted, and a core particle sorting algorithm is combined to optimize a placement sequence. According to the method, the quality and speed of layout planning are remarkably improved, and remarkable calculation acceleration is achieved through the rapid thermal model.

A computing system may include an access engine and a toolpath reordering engine. The access engine may be configured to access an original layer toolpath for slice of a 3D CAD object as well as a heat criticality measure for the original layer toolpath. The heat criticality measure may specify a heat impact for different points on the multiple toolpath segments of the original layer toolpath for the 3D printing of the physical part using the original layer toolpath. The toolpath reordering engine may be configured to reorder the multiple toolpath segments into a modified layer toolpath, and the modified layer toolpath may have a heat criticality measure with a lesser heat impact on the physical part than the heat criticality measure for the original layer toolpath.