177 results about "Non critical" patented technology

The invention provides an FPGA lookup table merging and decomposing method and device, equipment and a medium, and relates to the technical field of integrated circuit netlist optimization. The method comprises the following steps: acquiring a mapped LUT network of an FPGA; then, candidate pair identification is carried out on the LUT network, conflict-free LUT candidate pairs with the optimal weight are selected to be merged, an LUT connection relation graph is constructed after merging, and a key path is identified; on the basis of the LUT connection relation graph, decomposing the LUT on a non-critical path into smaller cascaded LUT units according to a LUT priority sequence; and dynamically updating the network topology structure and time sequence information of the LUT connection relation graph after each decomposition to complete optimization of the FPGA netlist. According to the method, the number of LUTs can be remarkably reduced and the resource utilization rate of the FPGA can be improved on the premise of ensuring that the delay of the critical path is not changed.

The invention relates to the technical field of electric digital data processing, in particular to a BIM building model construction method for building engineering design, which comprises the following steps: determining a newly added building area of each construction stage relative to the previous construction stage based on the change condition of a BIM building model of a building project in different construction stages; key areas and non-key areas are screened out from all newly-added building areas; based on the construction complexity of the key area, the criticality of the key area is determined, and then the adaptive LOD hierarchy of the key area is determined; allocating a set LOD layer not higher than the adaptive LOD level for the non-key area; and constructing a BIM building model based on the set LOD hierarchy allocated to the non-key areas of all the construction stages and the adaptive LOD hierarchy of the key areas. According to the method, the self-adaptive LOD levels are determined for different key areas, so that the calculation amount is reduced while the construction quality of the BIM building model is ensured.

The invention belongs to the technical field of artificial intelligence, and provides a deep learning backdoor attack detection method and device based on a non-critical feature dynamic threshold. The method comprises the following steps: inputting a to-be-detected image to a second image classification model to obtain a non-key feature map and a classification result of the to-be-detected image; the second image classification model comprises a feature extraction network; the screening module is used for separating a key feature map and a non-key feature map from the feature map of the to-be-detected image by using the mask; the classifier is used for carrying out classification processing on the key feature map of the to-be-detected image to obtain a classification result; calculating a non-key feature score based on the non-key feature map of the to-be-detected image, and if the non-key feature score reaches a preset score threshold, setting the abnormal mark of the to-be-detected image as abnormal; and outputting the abnormal mark and the classification result of the to-be-detected image. According to the method, extremely fine adjustment of an attacker on some pixels of the image can be identified, and the method can adapt to hidden attacks such as feature space disturbance.

The invention discloses an intelligent processing method and system for clinical test data, and relates to the field of data processing, and the method comprises the steps: recognizing a high flow disturbance stage, and starting an uploading channel adjustment process; executing flow limiting operation in the high flow disturbance stage; obtaining a load pressure evaluation result; and completing data supplementation. According to the method, under the condition of data uploading peak or resource shortage in the clinical test process, the whole-process regulation and control of the disturbance data, such as current-limiting protection, back compensation and abnormal filing, is realized by effectively identifying the high-flow disturbance stage, dynamically adjusting the data uploading behavior and intelligently distinguishing the processing priorities of the key data and the non-key data. The method has the advantages of being high in adaptability, accurate in resource scheduling and safe and reliable in data, solves the problems of system congestion, uploading failure or important data lag in the data uploading stage, and improves the overall operation stability and the data integrity guarantee capability of a clinical test data processing system.

The invention relates to the technical field of electrical fault detection, in particular to an online fault judgment and automatic detection method for a marine cable, which comprises the following steps of: acquiring ground leakage current and load current in operation, extracting an active component of the leakage current and analyzing the nonlinear correlation between the active component and the load current to realize online early warning of insulation degradation; when the ship is in a working condition of allowing short-time power failure of partial non-critical loops, positioning a fault branch based on a normalization result of the ground capacitance variation before and after the power failure of each branch and the cable length; after power supply is recovered, the fault type is judged by combining the leakage current fundamental wave and phase voltage phase difference and the harmonic distortion rate; during power-off maintenance, high-frequency carrier waves are injected, the position of a fault point is determined according to current attenuation abnormity, correction is carried out by combining temperature and vibration factors, and the detection accuracy and the positioning precision are improved.

The invention discloses an elastic scheduling method and system for a power grid based on a dynamic security domain, and relates to the technical field of elastic scheduling methods, and the method comprises the steps: determining the boundary of a non-key constraint based on an elastic adjustment factor, a power grid operation event corresponding to the power grid, and a plurality of constraints; the corresponding optimization event is determined according to the dynamic security domain, the boundary of the non-critical constraint and the real-time working data of the power grid, and the elastic scheduling system of the power grid is determined based on the optimization event, the real-time load data of the power grid and the corresponding real-time power supply data, so that the accuracy of the elastic scheduling system of the power grid is improved; and the elastic scheduling effect of the power grid is further improved. Therefore, the plurality of sub elastic scheduling projects are determined according to the identification of the elastic scheduling mode, the power optimization event of the power grid is determined based on the project content of each sub elastic scheduling project, the corresponding project position and the real-time load data of the power grid, and the accuracy of the online optimization logic of the power grid is ensured.

The invention relates to the technical field of laser cutting, in particular to a waterproof material laser cutting path optimization method based on deep learning, which comprises the following steps: acquiring basic size data, thickness change rate distribution data and preloading stress distribution data of a target test piece, and identifying key / non-key path sections through discretized grid units; cutting is conducted on the non-critical path section through basic cutting parameters, edge size data are monitored in real time, and the parameters are dynamically adjusted to form verification cutting parameters; screening a first confidence path segment based on stable cutting quality, and extracting verification parameters to construct a confidence collection data set; key cutting parameters are determined through parameter fluctuation analysis and correlation calculation, and a confidence data set is formed; cutting the key path section by adopting the parameter; and finally, establishing a deep learning model based on the historical confidence data set, and outputting basic cutting parameters of the new test piece. According to the method, self-adaptive optimization of cutting parameters and paths is achieved, and the cutting precision of the waterproof material test piece is remarkably improved.

The invention relates to the technical field of safety control of mining equipment, in particular to a mining intrinsic safety type multi-mode cooperative control system. The system comprises a main controller, a sensor module, a voice alarm module and an equipment execution unit. According to the invention, by setting a hierarchical protection shielding function, an effect of accurately preventing and controlling major safety risks is achieved, based on a dual judgment mechanism of a comprehensive risk coefficient and an operation risk index, an early warning-alarm-emergency response three-level protection system is systematically constructed, and when risks of equipment temperature abnormity, speed overrun and the like are detected, the system can give an alarm. The method comprises the following steps: firstly, triggering voice alarm and reserving emergency disposal time, and automatically shielding a non-key protection function when the risk is upgraded to a system-level emergency state, thereby avoiding economic loss caused by whole-line production halt, breaking through the limitation of once-by-once stopping of a traditional safety system, and remarkably improving the intrinsic safety level of mine production.

The invention provides a processor resource management method and a related device. The electronic device may identify critical threads, critical dependent threads, and non-critical threads. When the key thread and the key dependent thread are operated, the electronic device can dynamically adjust the frequency of the processor according to the task processing conditions of the key thread and the key dependent thread, and accurate supply of processor resources to the key thread and the key dependent thread is achieved. An electronic device may constrained provisioning of processor resources provided to non-critical threads. The method can improve the utilization efficiency of processor resources. According to the electronic equipment, the electric quantity of the electronic equipment can be saved while the experience of smoothly using the electronic equipment is provided for a user, the situation that the temperature of the electronic equipment is too fast is reduced, and balance between equipment performance smoothness and the requirements of long endurance and rapid cooling is achieved.

The embodiment of the present specification provides a task processing model training method, a task processing method and a request processing method based on the task processing model. The task processing model training method comprises the following steps: obtaining a sample set and a test result of a sample task; determining a key processing unit corresponding to the sample task from a plurality of processing units according to the test result, wherein the influence degree of the key processing unit on the test result is greater than the influence degree of a non-key processing unit on the test result, and the non-key processing unit is a processing unit other than the key processing unit in the task processing model; and adjusting parameters of the key processing unit according to the sample set to obtain a trained task processing model. The key processing unit is determined before the model training, and only the parameters of the key processing unit corresponding to the sample task are adjusted, which reduces the mutual interference between different tasks, effectively solves the performance conflict problem, and improves the multi-task processing performance of the task processing model.

The invention belongs to the field of artificial intelligence platform development, operation and maintenance, and particularly relates to an artificial intelligence platform task management method and system. The method specifically comprises the steps of creating a Pod corresponding to a training task of an artificial intelligence platform through K8s; the created Pod comprises a resource demand of a training task; the K8s is utilized to allocate nodes meeting resource requirements to the Pod according to the resource requirements of the training tasks so as to execute the training tasks corresponding to the Pod; reducing the upper limit value of the resource demand of the Pod corresponding to the non-key training task in a set resource consumption peak period; and increasing the upper limit value of the resource demand of the Pod corresponding to the non-key training task in the set resource consumption valley period. In this way, it can be ensured that the key tasks are executed preferentially in the time period of resource shortage, completion of the non-key training tasks is accelerated in the time period of resource sufficiency, and therefore the completion progress of the training tasks of the artificial intelligence platform is optimized on the whole.

The invention relates to a method for manufacturing a donor wafer comprising the following steps: a) supplying a donor substrate made of a single-crystal material, said donor substrate having been recycled at least once, after a transfer of a thin layer of said donor substrate onto a receiving substrate so as to form a composite structure, said composite structure comprising a surface layer formed in whole or in part by the thin layer and disposed on the receiving substrate, b) checking the quality of the composite structure by inspecting a free surface of the surface layer to detect defects present on and / or in the surface layer, said defects then being classified into a first category, considered non-critical, or into a second category, considered critical.c) the selection of the donor substrate if the density of defects classified in the second category during step b) is less than a predetermined density, d) the assembly of the donor substrate selected in step c) onto a support substrate, by bonding, to form the donor wafer. The invention also relates to a donor wafer comprising a single-crystal silicon carbide donor substrate, having a thickness between 100 μm and 300 μm, disposed on a support substrate, and characterized in that a free face of the donor substrate has fewer than 0.5 micro-hole or inclusion defects / cm². Figure to be published with the abstract: No figure.

This invention relates to the field of communication technology, providing a communication method and device for a charging pile and its control board. The method includes: when a data transmission request event is detected, determining the target communication link based on an address mapping table and transmitting data; if reliable transmission is required for the data, storing a copy of the data in a pending confirmation queue and starting a retransmission timer; when a retransmission timer trigger event is detected, checking the unacknowledged data in the pending confirmation queue and retransmitting it according to a pre-set retransmission strategy for the communication link; and when an ACK response message is received, matching and removing the corresponding data from the pending confirmation queue. The retransmission strategy is dynamically configured based on the communication link medium type and data priority. This invention, by configuring differentiated reliable transmission mechanisms for different medium types and different priorities, ensures the reliability of critical instruction transmission while reducing bandwidth consumption by non-critical data.

Methods and systems for improved impact detection in a device, such as a head-wearable device, are disclosed. The methods and systems relate to accurately detecting impact events such as falls or vehicular crashes to reduce false positives and enable a swift, appropriate emergency response. A method includes detecting use of the device by a user, for instance, confirming that a head-wearable device is being worn. The method further includes detecting a change in at least one of an orientation or a movement of the device within a threshold time period. An impact is then determined based on the confirmed use of the device and the detected change in its orientation or movement. This approach distinguishes genuine emergencies from non-critical events like dropping the device, thereby improving accuracy and conserving power by intelligently activating additional sensors only when a potential impact is detected.

Systems, methods, and other embodiments described herein relate to safely activating a fuel cell (FC) within a generator. In one embodiment, a method includes initiating a test for sensitive systems of a generator during a standby status using backup power including a battery. The method also includes powering an FC and a direct current (DC) converter on a first bus within the generator by increasing an operational voltage on a second bus using the battery. The method also includes, upon successfully completing the test and powering the FC and the DC converter, energizing a load inverter by switching power flow from the battery to the FC after completing a non-critical sequence that controls support systems of the generator for a generating status, wherein the DC converter stabilizes voltage energy between the FC directly connected on the first bus above a minimum voltage and output energy to the second bus.

This invention discloses an optimization method and system based on mechanism importance and defect merging. The method includes: collecting multi-source evidence of bridge defects; determining the occurrence mechanism and impact dimensions of various defects through coupled defect model matching; classifying defects into critical defects affecting structural safety or causing systemic failure and non-critical defects affecting serviceability or durability based on the risk level of the upstream defect diagnosis module; establishing a defect importance classification list and clarifying the classification basis and judgment criteria; obtaining the preliminary defect assessment scale and the risk level and development trend given by the upstream module; maintaining the preliminary scale for critical defects and adjusting the optimization scale of non-critical defects according to a preset calibration rule set; for calibrated non-critical defects, if they belong to a preset minor defect merging pair, merging them into one item and determining their representative scale and contribution order. This ensures bridge operation safety and improves maintenance economy.

The invention discloses a state evaluation method and system of a photovoltaic power generation assembly, and relates to the technical field of state evaluation of the photovoltaic power generation assembly, and the method comprises the steps: determining whether there is only one crack type feature in a target key region of a target assembly according to the area of a crack coverage region and the concentration degree of cracks in the coverage region if there is only one crack type feature in the target key region of the target assembly; evaluating the induced fault risk value of the target component, if no less than one crack category feature exists and if a crack crossing phenomenon exists, evaluating the induced fault risk value of the target component according to the crack crossing number, and comparing the obtained first or second risk evaluation value with a preset key risk early warning threshold value, or determining the risk value of the induced fault of the target component according to the preset key risk early warning threshold value. And comparing the third risk assessment value or the fourth risk assessment value to which the target non-key area of the target component belongs with a preset non-key risk early warning threshold, and performing different levels of risk early warning. According to the state evaluation method and system for the photovoltaic power generation assembly, the risk early warning efficiency can be improved.

A computer-implemented method includes training a machine-learning model, using a training dataset that distinguishes between critical systems and non-critical systems, to classify a particular comput

The method, the control device (7), the computer program product (11) and the fuel injection device (1) according to the invention are designed to operate a fuel injector (4) which has an injector valve (4a) and is associated with a combustion chamber (3a) of the internal combustion engine (2), in particular within a critical operating range of the fuel injector (4). For this purpose, an electronic control device (7) associated with the fuel injector (4) first applies a test control signal to the injector valve (4a) during a non-critical operating interval and receives a corresponding response signal from the fuel injector (4) and determines therefrom an opening duration deviation of the injector valve (4a). Control parameters for controlling the fuel injector (4) to inject fuel in a critical operating interval of the fuel injector (4) are determined by means of a transfer function as a function of a specified target fuel injection amount and a determined opening duration deviation of the injector valve (4a), and a current pressure and / or a current fuel temperature, and a corresponding control signal is output to drive the fuel injector (4). In this way, a nominal specified fuel injection quantity and optionally a desired injection starting time can be achieved with high accuracy even in a critical operating range of the fuel injector (4).

The invention discloses a power plant unit control method and system based on edge calculation, and relates to the technical field of edge calculation. The method comprises the following steps: acquiring and processing a first data set in real time by utilizing an edge computing node; when the key time period is identified, triggering a corresponding dynamic adjustment mechanism; the technical key points are as follows: on the basis of an originally divided key period and an originally divided non-key period, the period is divided twice to obtain a calibrated period and a non-calibrated period, different constraint condition correction factors are adopted to adjust an originally determined probability factor, so that a control strategy is more targeted, and by constructing a parameter adjustment model, the accuracy of the control strategy is improved. The time difference adjusted in advance is output, and the control is dynamically adjusted according to the time difference, so that a preset management mode is realized, the accuracy of regulation and control is realized, much inconvenience caused by instant response is avoided, the problem that a traditional single control strategy is difficult to meet diversified working requirements is solved, and the flexibility and adaptability of the control strategy are improved.

The utility model relates to the technical field of forming dies, in particular to a forming die with a guide structure, which comprises an upper die plate, the bottom of the upper die plate is in threaded connection with the top of an upper die, and a lower die is arranged at the bottom of the upper die. Guide plates are arranged on the outer walls of the left side and the right side of the upper mold; a pressing groove is formed in the bottom of the upper mold; and guide grooves are formed in the outer walls of the left side and the right side of the lower mold. According to the device, the upper die plate and the upper die are rapidly and accurately assembled according to the design of corresponding screw holes, the connection firmness is enhanced, disassembly and maintenance are convenient, the T-shaped guide plate is inserted into the guide groove to limit transverse deviation and automatically align, and the part forming precision and the service life of the die are guaranteed. The pressing groove and the curved surface of the lower die are attached to the outline of a part, plate materials are guided to deform evenly, cracks are reduced, and shape precision is ensured. The weight of the upper die weight reducing groove is reduced by removing non-critical materials, the equipment load is reduced, the operation safety and the forming stability are improved, and material utilization and function balance are optimized.

Techniques are disclosed for a fault monitor to determine that no fault has been lost. The fault monitor may distinguish between critical and non-critical faults without the risk of missing a critical fault. The fault monitor may detect a new fault event from a fault source and may set a fault overrun flag when the new fault event is detected while an existing fault event from the fault source is being processed. When the fault monitor receives a fault acknowledgement to indicate completion of processing of the existing fault event, the fault monitor may generate a request to process the new fault event as an overrun fault based on the fault overrun flag is set. The fault monitor may generate a request to process the new fault event as a non-overrun fault when the new fault event is detected without an existing fault event from the fault source being processed.

The application discloses a kind of translation lookaside buffer table entry processing method, system, medium and product, applied to processor technical field, comprising: when process switching, determine the non-critical table entry corresponding to the process before switching in translation lookaside buffer, obtain target non-critical table entry;The target non-critical table entry is stored in the first cache corresponding to processor core, and the target non-critical table entry is deleted from the translation lookaside buffer;The recovery deadline corresponding to the process before switching issued by operating system is acquired;Based on the recovery deadline, before switching back to the process before switching, the target non-critical table entry is written back to the translation lookaside buffer from the first cache.This way, in the case of guaranteeing the process before switching fast recovery, the hit rate of translation lookaside buffer table entry can be improved, and then the processor performance is improved.

This application relates to digital twin 3D modeling, specifically to a method and device for constructing a digital twin waterway scene using electronic navigation charts. The method includes: analyzing the historical trend of water depth values ​​at each depth point to determine the inherent terrain feature values ​​of each depth point set at the current time; determining the terrain ruggedness feature value by fitting the last data point in the temporal sequence of the trend components of all depth points in each depth point set at the current time with its fitted value and the fitting error; and determining interpolation feature values ​​by combining the inherent terrain feature values ​​to classify all depth points into critical and non-critical points; and performing interpolation processing on the water depth values ​​of all critical and non-critical points using different interpolation methods respectively. This application solves the problem that a single interpolation method cannot simultaneously ensure the realism and real-time performance of the model construction, thus improving the realism and real-time performance of the digital twin waterway scene model.

Disclosed are system and methods of dynamically balancing power and performance of a processing unit. The method includes receiving workloads to be processed by the processing unit; classifying, using a recurrent neural network, the workloads according to expected resources to be expended by different functional blocks; identifying a critical path workload based on the classification of workloads; determining a temperature-independent operating frequency for the critical path workload; determining a temperature-dependent operating frequency for the critical path workload based on a junction temperature of the processing unit; adjusting the temperature-dependent operating frequency to a target operating frequency having a value within a predetermined difference from the temperature-independent operating frequency by actively cooling the processing unit; setting a critical path supply voltage based on the target operating frequency; and setting a non-critical path voltage for remaining ones of the workloads at a value less than the critical path supply voltage.

A memory controller is provided and configured to control access to a memory with a memory access bandwidth by a plurality of processing components of a data processing system. The memory controller comprises a control unit configured to allocate a respective fraction of the memory access bandwidth to each of the plurality of processing components for accessing the memory by a plurality of memory access operations. Moreover, the memory controller comprises a monitoring unit configured to obtain timing information about a duration of a respective memory access operation by each of the plurality of processing components via a bus of the data processing system. The control unit is further configured to adjust, for one or more of the plurality of processing components, the respective fraction of the memory access bandwidth based on the timing information obtained by the monitoring unit.

Disclosed in embodiments of the present disclosure are a method and apparatus for applying a large language model. A specific implementation of the method comprises: inputting an original question to be queried and a preset prompt into a large language model, and outputting a first question, wherein the prompt is used for instructing the large language model to extract, from the original question, a question reflecting a real intention; deleting non-key information from the first question to obtain a second question; and inputting the second question and a preset security prompt into the large language model, and outputting an answer, wherein the security prompt is used for indicating that the large language model needs to comply with a security policy when answering a question. The implementation can effectively defend against jailbreak attacks having deceptive and implicit intentions.

The invention relates to the technical field of collaborative scheduling management, in particular to a tide and weather linkage scheduling method for port operation collaboration, which comprises the following steps of: setting a benchmark, creating a virtual time axis, the initial state of the virtual time axis being a standard uniform elapsed time axis; marking an effective operation time window on the virtual time axis according to forecast tide data; performing correction according to the forecast meteorological data to generate a deformation time axis; calculating a starting time point and duration of each operation link on a deformation time axis, and reversely deducing a collaborative operation instruction sequence to form a final scheduling scheme; and dynamically adjusting the deformation time axis, and triggering rolling optimization of the final scheduling scheme. According to the method, it is guaranteed that the scheduling scheme is always unfolded around the core target of safe departure, time consumption of non-critical paths is reduced to the maximum extent, and the scheduling failure problem that in a traditional method, scheduling is completed, but departure fails is avoided.

This invention relates to a method and system for real-time monitoring of the additive manufacturing (AM) building process of a three-dimensional (3D) product generated by a 3D building device. The building process comprises alternately distributing a material on a substrate by a coating mechanism and forming a layer by fusing a portion of the material according to a planned geometry. An imaging device captures image data for each build layer, including a pair of images representing the distributed material prior to layer formation and the corresponding build layer after formation. A processor processes the image data by performing a localized-image evaluation to detect local surface defects before and after layer formation and determine a localized-image failure metric, and / or by performing a whole-image evaluation using statistical learning models to determine a whole-image failure metric. Based on the comparison of the failure metrics with reference values, each layer is categorized as critical or non-critical. The method further includes determining a forward-looking failure risk indicator using the evaluation results and a defined number of most recent, to predict whether the building process is trending toward failure.