51 results about "Elimination method" patented technology

The invention discloses an invalid code automatic elimination method and device, a medium and a product, and the method comprises the steps: carrying out compiling analysis processing on each program file in a to-be-processed target program item, generating a syntax tree model, and testing the compiled program file to obtain a program test result; unused variables in all the variables are identified through a syntax tree model, and uncovered code lines are positioned in combination with a test result; analyzing logic judgment conditions of uncovered code lines based on the syntax tree structure, and judging the code lines which cannot be triggered as unreachable codes; unused variables and unreachable codes in the program files are removed, new program items are obtained, and when the new program items are correctly compiled and executed, the new program items are used for replacing the target program items. Program redundancy is reduced, and execution efficiency is improved.

The invention discloses a redundancy constraint identification and elimination method for large-scale power market clearing simulation, and the method comprises the steps: building a safety constraint unit combination and safety constraint economic dispatching optimization model in which a tracking unit, a new energy unit and an energy storage power station jointly participate in market clearing; factors such as unit self-constraint, new energy randomness and energy storage charge-discharge balance are comprehensively considered; according to a large number of line power flow constraints involved in the clearing process, the maximum possible transmission power of a power transmission line is calculated, the conversion relation between unit power and load requirements on the line is combined, and redundancy of safety constraints is rapidly recognized and screened through an analysis method, so that efficient elimination is achieved; and solving the node marginal electricity price based on the solved unit combination start-stop scheme result and the declared unit output plan. The method is more suitable for a large-scale market environment, can effectively reduce the consumption of computing resources, improves the real-time performance of market clearing, and optimizes the operation management of a power transmission network.

The invention discloses a data processing method, an idle stroke eliminating method, an anti-collision method and related equipment thereof. The idle stroke eliminating method comprises the following steps: acquiring a real-time physical quantity of a main shaft when a machine tool machines a first workpiece; processing the real-time physical quantity according to a preset rule to determine a first threshold value; the first threshold value serves as a boundary threshold value when the second workpiece is machined; compared with the problem that a fixed boundary threshold value is determined depending on manual experience at present, the method comprises the steps that the real-time physical quantity of a main shaft is obtained when a machine tool machines a first workpiece; processing the real-time physical quantity according to a preset rule to determine a first threshold value; the first threshold value is used as the boundary threshold value when the second workpiece is machined, so that the boundary threshold value when the machining tool machines can be adjusted in real time.

The invention provides a cache elimination method and device, a storage medium and computer equipment, and relates to the technical field of computers.The method comprises the steps that a weight label is generated for each cache data item in a cache space, and the weight label comprises a service priority field and an access frequency field; for each cache data item, when the cache data item is accessed, calculating the current frequency score of the cache data item according to the historical frequency score and the current access increment of the cache data item, and updating the access frequency field of the cache data item; and when the cache space is full, determining a target elimination data set according to the weight tag of each cache data item, and eliminating the cache data items in the target elimination data set from the cache space. In this way, the problem of cache breakdown can be effectively relieved, and therefore the technical defect that in the prior art, due to the fact that hot data are eliminated, the system performance is reduced is overcome.

The invention discloses an adaptive content eliminator and elimination method for an image generation large model, and the method comprises the steps: constructing an elimination space and a reserved space according to a target content, the elimination space corresponding to semantic information needing to be removed, and the reserved space corresponding to semantic information needing to be reserved; converting a text prompt input by a user into semantic representation, and projecting the semantic representation into the elimination space and the reserved space token by token to obtain a reserved component of the reserved space; measuring a semantic distance between the semantic representation of the current text prompt and the semantic representation of the target content, and calculating a correlation degree; according to the correlation degree result, the reserved space is reconstructed, and the direction of the reserved component is corrected. The method does not need to consume a large amount of computing resources to finely adjust the model, and does not introduce additional parameters. And when the single or multiple target contents are successfully removed, the influence on the generation of irrelevant contents can be ignored, so that the content elimination performance of the generative model is remarkably improved.

The invention discloses a cache data elimination method, electronic equipment, a storage medium and a program product, and relates to the technical field of computers, and the method comprises the following steps: based on the data type of to-be-eliminated data in a to-be-eliminated node, inserting the to-be-eliminated node into a corresponding elimination chain table; if the use space of the cache is greater than a preset use space threshold value, determining a value influence factor of cache data of a target data type and a current cache use space for any elimination chain table based on the target data type corresponding to the elimination chain table, and further determining an elimination weight of the elimination chain table; eliminating the target elimination node in the elimination chain table; the larger the value influence factor and the current cache use space are, the larger the elimination weight is; the smaller the value influence factor is, the higher the cache income is; the greater the elimination weight is, the more the target elimination nodes are. The technical problem that the cache data elimination mode in the related technology increases the data access delay is solved, and the technical effect of reducing the data access delay is achieved.

The invention discloses a large model illusion elimination method and device, equipment and a storage medium, and relates to the technical field of artificial intelligence natural language processing. Carrying out atomization disassembly on the initial generation content to obtain an atomic fact set; performing internal knowledge base semantic matching and networking retrieval cross validation on each atomic fact in the atomic fact set to obtain an evidence coverage rate; when the evidence coverage rate is lower than a preset coverage rate threshold value, generating an evidence vacancy report; performing differentiation rewriting on the initial generation content based on an evidence vacancy report to obtain target generation content, performing evidence coverage rate quantitative checking on the generation content through atomization disassembly and an internal and external double-source cross validation mechanism, blocking illusion cascade in the early stage of reasoning, and feeding back a verification result for iteration rewriting, so as to obtain the target generation content. And high-confidence content subjected to dominant verification is obtained.

The invention discloses a violation elimination method and device based on a cost function, equipment and a medium, and relates to the field of chip manufacturing, and the method comprises the steps: for a current chip design pattern after optical proximity correction, traversing each edge in the current chip design pattern, and determining a plurality of groups of edges based on the traversed current edge and a plurality of surrounding candidate edges; determining a corresponding cost function based on the mask rules to which the plurality of groups of edges respectively belong so as to calculate the cost value of the corresponding group of edges; determining a violation from the plurality of sets of edges based on the cost value; and moving the violation position to update the current chip design graph, and returning to the step of traversing each edge in the current chip design graph until no violation exists in the plurality of groups of edges. According to the method and the device, the violation is determined from the plurality of groups of edges based on the cost value of each group of edges by searching the whole solution space, and the whole solution space is searched again after the violation is eliminated until all the violations are eliminated, so that all the violations can be eliminated more easily through a larger search space.

The invention relates to a method for automatically eliminating a video object. The method comprises the following steps of: acquiring a video to be processed, identifying and determining a target object needing to be eliminated, generating a segmentation mask by using an automatic tracking technology, segmenting the target object frame by frame and determining the boundary of the target object, and then automatically eliminating the target object in each frame of video. The video background is repaired through the generated optical flow information, the time sequence consistency of the video is maintained, a blank area generated by object elimination is further filled by using a feature propagation technology, and finally a target video with high quality and good consistency is generated. According to the method, the video processing efficiency is remarkably improved, the error rate in manual operation is reduced, the visual effect and the overall quality after video processing are ensured, and the method is suitable for video editing tasks needing high precision and high efficiency.

The application claims a time-oriented RDF quad model for MongoDB storage and a redundant attribute elimination method, which comprises the following steps: step 1, constructing a time-oriented RDF quad data model, taking time information as a new tuple of RDF triple extension, expanding the triple into a quad model containing time information, and defining the expanded time-oriented RDF quad data model; step 2, storing time-oriented RDF quad instance data in a document-oriented non-relational MongoDB database; step 3, using a weighted-based design algorithm to find out the instance attributes with higher redundancy from the time-oriented RDF quad data in the form of text stored in the MongoDB database, and eliminating these redundant instance attributes in the ontology library that has been constructed. The RDF quad-based query model designed in the application has a more significant effect on resources with time characteristics, and at the same time solves the problem of ontology ambiguity caused by repetitive instance attributes.

The invention belongs to the technical field of artificial intelligence, and relates to a semantic level concept elimination method and system based on domain concept mining. The method comprises the following steps of: mining an associated concept by adding random perturbation in embedding in a neighborhood space for eliminating a concept; mapping the elimination concept and the mined association concept to an anchor concept of a diffusion model, and realizing concept elimination by finely tuning the diffusion model; and finding a reserved concept irrelevant to the elimination concept through out-of-neighborhood concept regularization, and reserving the reserved concept. According to the method, the associated concept is guided to the anchor concept through a neighbor concept mining mechanism, and specificity retention is performed by adopting out-of-neighborhood concept regularization, so that the negative influence of the associated concept on the unassociated concept is compensated and eliminated. The method is superior to an existing concept elimination method in the aspects of elimination effectiveness, specificity and universality.

The present invention discloses an echo cancellation method based on adaptive decorrelation and variable step-size proportional M estimation, comprising the following steps: A. signal acquisition; B. calculating the decorrelation coefficient vector #imgabs0# according to the speech signal sampling and the expected signal sampling results; C. calculating the decorrelation input vector x D (n) and the expected signal d after decorrelation D (n); D. Calculate the output y(n) of the adaptive filter and the output y of the adaptive filter after decorrelation D (n); E, calculate the error signal e(n) and the error signal e after decorrelation D (n); F, calculate the proportional matrix G(n); G, calculate the error signal #imgabs1# after robust decorrelation for iteration in M ​​estimation; H, calculate the step size parameter μ(n); I, update the tap weight vector of the adaptive filter and enter the processing at the next moment. In addition to having good robustness to impulse noise, the present invention also has the advantages of fast convergence speed and good steady-state error.

The invention relates to the field of data processing, in particular to a character ambiguity elimination method and system based on artificial intelligence. The method comprises the following steps: constructing a lexicon, and dividing all reference words into strong correlation words and weak correlation words of target words; removing unintentional words in the answer sentences to obtain effective answer sentences, calculating a distribution coefficient of any word in the effective answer sentences, and constructing word vectors of the effective answer sentences according to the distribution coefficient; clustering the word vectors of all the effective answers to obtain a plurality of clusters, traversing any cluster to obtain reciprocals of the number of the strongly related words in the marked question corresponding to each effective answer, and taking a mean value of all the reciprocals as an ambiguous value of the cluster; and traversing to obtain an ambiguity value of each cluster, constructing an ambiguity value sequence of the contact word to the target word, and calculating ambiguity evaluation to complete ambiguity elimination. According to the technical scheme, the accuracy of text understanding can be improved, and recognition errors caused by character ambiguity are reduced.

The invention provides a TSN redundant message elimination method and device, and the method comprises the steps: constructing an HASH + CAM flow table structure which takes a triple HASH table as a main part and a CAM table as an auxiliary part, and pre-storing a Key value of a TSN message flow and a corresponding flow ID number in the HASH + CAM flow table structure; obtaining a TSN message to be judged, extracting key information of the TSN message, and generating a Key type and a Key value based on the key information; querying in the triple HASH table according to the Key type and the Key value, and outputting a stream ID number corresponding to the Key value when the Key type and the Key value are hit; if the query of the triple HASH table is not hit, performing query in the CAM table according to the Key value, and outputting a stream ID number corresponding to the Key value when the query is hit; and confirming whether the TSN message is a redundant message or not according to the stream ID number, and discarding the redundant message after confirmation. Through efficient HASH table and CAM table design and optimized table item storage and access strategies, redundancy elimination of large-flow TSN messages under low clock frequency is realized.

The invention discloses a self-adaptive backlash elimination method, device and equipment based on dual-motor coaxial driving and a storage medium, and relates to the technical field of electromechanical control, and the method comprises the steps: controlling a first motor and a second motor to carry out calibration when an output shaft of a control gearbox is fixed, and obtaining a backlash elimination value; based on the backlash elimination value and the user target position, motion track planning is obtained; and when the first motor and the second motor are controlled to move according to the motion track plan, adjusting the backlash elimination value according to the current value in the backlash elimination stage and the current value in the target motion stage, so that the adjusted backlash elimination value gradually reaches the actual backlash value. The technical problems that due to the fact that the tooth space value changes along with operation abrasion, the preset compensation value loses efficacy, the double motors move asynchronously, and the transmission shaft generates torsion stress are solved, the tooth space value is eliminated, the tooth space value can automatically change along with the actual tooth space value, and therefore position synchronization and torque balance are maintained for a long time in a double-motor coaxial driving system.

The invention provides a method, a device and equipment for intelligently eliminating keyboard key conflicts. The method comprises the following steps: acquiring a key event signal sequence when a user uses keyboard keys; constructing a conflict situation map of a plurality of key pressing states based on the key event signal sequence; monitoring a foreground application program which is currently in a focus state, acquiring context semantic information of the foreground application program, identifying an input membership degree, belonging to real intention input of a user, of each key position in the conflict situation map under the current context based on the context semantic information, and further generating an input membership degree distribution map of a current keyboard; and according to the input membership distribution diagram, redistributing the signal output priority of each key, generating a virtual signal output queue with an adjustable time sequence, and further eliminating the conflict of the keyboard keys based on the virtual signal output queue. According to the technical scheme provided by the invention, the key priority can be identified when a plurality of key signals are overlapped.

The present invention discloses an adaptive content eliminator and an elimination method for an image generation large model. The elimination method includes: constructing an elimination space and a retention space according to the target content, where the elimination space corresponds to the semantic information to be removed, and the retention space corresponds to the semantic information to be retained; converting the text prompt input by the user into a semantic representation and projecting it token by token into the elimination space and the retention space to obtain the retention components in the retention space; measuring the semantic distance between the semantic representation of the current text prompt and the semantic representation of the target content, and calculating the correlation degree; and reconstructing the retention space and correcting the direction of the retention components according to the correlation degree result. The present invention does not require a large amount of computing resources to fine-tune the model and does not introduce additional parameters. While successfully removing single or multiple target contents, the impact on the generation of irrelevant contents is negligible, significantly improving the content elimination performance of the generation model.

The invention belongs to the technical field of federated learning, and particularly relates to a backdoor elimination method based on a client in federated learning. Aiming at the problems that federal learning is easily attacked by an inter-wheel backdoor due to open participation, the existing screening type aggregation is invalid, and the precision of a relieving technology is reduced, backdoor elimination is realized by adopting client side periodical alternating layer resetting, undifferentiated forgetting and knowledge distillation collaboration, and the method comprises the following steps: firstly, obtaining a global model, a local data set, a historical model and undifferentiated forgetting parameters; when a reset condition is met, the specified layer parameters are reset according to layer group parity; secondly, performing undifferentiated forgetting training on local data to decouple back door association, and performing knowledge distillation by taking the historical model as a teacher and the model after forgetting as a student to recover the performance of a main task; and finally, outputting the purification model for local training. According to the method, server or trigger priori does not need to be modified, the cross-round attack success rate can be greatly reduced, the main task precision is not lost, and the federal learning robustness is enhanced.

The present invention provides an operator elimination method, system, electronic device and storage medium for a neural network, and relates to the technical field of deep learning. It includes: determining a target subgraph structure with operator elimination requirements in a neural network model; determining an operator in an eliminateable state in the target subgraph structure as a target elimination operator; traversing all subgraph structures until all target subgraph structures with operator elimination requirements and corresponding target elimination operators are determined; calculating the buffer memory sum of the input and output of the target elimination operator; determining the target cache area of ​​the target elimination operator according to the buffer memory sum, so as to allocate the target elimination operator to the target cache area, and configuring the output of all target elimination operators with the same cache address; determining the jump offset of the target elimination operator; compiling the target elimination operator based on the jump offset to complete operator elimination. The present invention can realize operator elimination and reduce the reasoning time of the neural network.

The application provides a multi-target ambiguity resolution and redundancy reduction method based on a multi-ary tree recursive model, comprising the following steps: step 1, according to a device repetition period parameter, a residual difference table is established for each two relatively prime repetition periods; step 2, after an apparent distance difference is obtained according to an apparent distance and a table is looked up to obtain an ambiguity, a target real distance is determined; step 3, according to the residual difference table, a matrix type staggered index two-dimensional table is established for a target index, a distance true value and a related repetition frequency index; and step 4, according to the established matrix type staggered index two-dimensional table, a method for obtaining multi-visibility of a target based on a multi-ary tree recursive model is adopted to perform a high-to-low fast search for multi-visibility of a target for all distance unit targets of each repetition period until the target is determined. The application effectively reduces redundancy of multi-target ambiguity resolution and improves processing speed.

The application provides a kind of active elimination method of DC ripple of wind power cluster sending-out system through diode dispersion rectification. In view of the strong time-varying characteristics of the harmonic amplitude and phase of the DC side of the diode rectification system changing with the output of the wind turbine and the reference phase of the AC side, which leads to the problem that the ripple suppression effect of the traditional fixed uniform interleaved angle control is limited, the method proposes an active elimination mechanism of DC ripple based on the negative gradient algorithm of the central controller. The central controller extracts the 12th harmonic amplitude and phase information of each DC branch current in real time, uses the negative gradient algorithm to minimize the total amplitude after the superposition of the 12th harmonic, dynamically adjusts the reference phase of the sinusoidal AC voltage of each wind turbine grid side, compensates the amplitude difference through active phase regulation, and realizes the closed cancellation of unequal amplitude harmonics under the condition of output imbalance. The application does not need to increase hardware, effectively suppresses the DC sending-out current ripple, significantly reduces the size and cost of the filter device, and is suitable for space-limited scenarios such as offshore wind power.

The invention discloses a multi-agent system conformity diagnosis and elimination method, a product, a medium and equipment. The method comprises the following steps: firstly, obtaining original questions and real answers of a logic analysis reasoning task and a language context understanding task, and formulating five interaction protocols for carrying out a conformity test on a multi-agent system; three evaluation indexes are designed to obtain a conformity diagnosis result of the multi-agent system; an independent personality enhancement mechanism of the multi-agent system is introduced to modify an original system cue word, an reflection mechanism of the multi-agent system is introduced to design an reflection mechanism cue word, and then the popularity of the multi-agent system is eliminated through the modified system cue word and the reflection mechanism cue word. The two prompt-based elimination technologies respectively correspond to enhanced independence before decision making of the intelligent agent and double check and reflection processes after decision making, and parameters of a large language model do not need to be finely adjusted.

The invention discloses a DC component digital elimination system and elimination method.The elimination system comprises a photoelectric detection module, an elimination module, a digital comparison output module and a control module, and the control module is used for closing an emitting element after elimination is triggered, controlling a first elimination digital-to-analog converter to traverse from the highest weight elimination gear and controlling a second elimination digital-to-analog converter to traverse from the highest weight elimination gear; determining a first elimination gear combination according to a comparison result of the current elimination gear, starting a transmitting element in a subsequent sampling period, controlling a first elimination digital-to-analog converter to counteract an ambient light signal, and controlling a second elimination digital-to-analog converter to traverse from the highest weight elimination gear when executing to a preset sampling gap, and determining the second elimination gear combination according to the comparison result of the current elimination gear to offset the direct current component in the reflected light signal, thereby dynamically determining the elimination gear combination in a gear-by-gear manner, effectively eliminating the direct current component in the ambient light signal and the reflected light signal, and ensuring the accuracy and reliability of the detection result.

The invention provides a large model refusal response elimination method and device, a storage medium and electronic equipment, and relates to the field of large language models. The electronic equipment obtains a mistaken rejection sample set and a conventional sample set; obtaining a first activation vector set when the to-be-processed model processes the false rejection sample set and a second activation vector set when the to-be-processed model processes the conventional sample set; determining an answer rejection network layer from the to-be-processed model according to the difference between the first activation vector set and the second activation vector set; and performing intervention on the answer rejection network layer, so that the to-be-processed model can normally process the harmless problem which is mistakenly rejected before intervention. Thus, by comparing the internal activation modes of the model for harmless but rejected questions and normally answerable questions, the rejection network layers dominating rejection behaviors are identified, and directional intervention is performed for the rejection network layers, so that the excessive defense response of the model can be weakened on the premise of minimizing the influence on other functions.

The objective is to obtain a biological elimination device that can prevent the target organism from becoming accustomed to a specific elimination method, depending on the number of times the target organism has invaded. [Solution] The system includes: an acquisition unit 7 that acquires detection information indicating that an organism to be eliminated 3 has entered the monitoring area 12 from a detection unit 5 that detects when an organism to be eliminated 3 has entered the monitoring area 12; and a control unit 11a that controls the elimination execution unit 6, which performs elimination to prevent the organism to be eliminated 3 from entering an intrusion-prohibited area, to change the elimination means executed by the elimination execution unit 6 to a different elimination means than the one executed when the cumulative number of eliminations was less than the predetermined number, if the cumulative number of eliminations, which is the number of times the organism to be eliminated 3 has entered the monitoring area 12, is greater than or equal to a predetermined number, based on the detection information.

The application provides a kind of self-interference continuous elimination method for the cooperative exploration of multiple unmanned ships in unknown areas, belonging to the technical field of autonomous exploration, autonomous positioning and map construction of multiple unmanned ships. To solve the self-interference problem that occurs when multiple unmanned ships cooperatively explore unknown areas, a coordinate relationship established by SLAM and multiple unmanned ship map fusion is used in combination with the iterative closest point registration algorithm to propose a laser radar point cloud method for continuously processing dynamic obstacles, which eliminates the self-interference of multiple unmanned ships in the cooperative exploration of unknown areas. In addition, for the laser radar point cloud data generated by dynamic obstacles, random points are formed in the map, which leads to poor subsequent multi-map fusion results and ultimately fails the exploration task. Through the coordinate system conversion relationship between unmanned ships, the application calculates the occurrence time of the self-interference problem and the initial position of the dynamic obstacle, and uses the iterative closest point to continuously track the position of the dynamic obstacle, achieving the purpose of continuously eliminating the influence of self-interference.

This invention discloses a method, product, medium, and device for diagnosing and eliminating conformity in multi-agent systems. The invention first obtains the original questions and true answers for logical reasoning and language context understanding tasks, and formulates five interaction protocols for testing the conformity of the multi-agent system. Three evaluation indicators are designed to obtain the diagnostic results of the conformity of the multi-agent system. An independent personality enhancement mechanism is introduced to modify the original system prompts, and a reflection mechanism is introduced to design reflection mechanism prompts. The modified system prompts and reflection mechanism prompts then eliminate the conformity of the multi-agent system. The two prompt-based elimination techniques of this invention correspond to the enhancement of independence before the agent makes a decision and the dual checking and reflection process after the decision, respectively, and do not require fine-tuning the parameters of a large language model.

The application discloses a multi-modal large model structured pruning and parameter redundancy elimination method, system, device and medium, comprising: initializing the structure of the multi-modal large model; dividing the pruning process into a far-end stage and a near-end stage; constructing a cross-modal perception structured parameter group based on the importance of low-rank matrix gradient estimation parameters, and calculating the group importance of each structured parameter group according to the importance of the parameters; arranging the structured parameter groups in ascending order of group importance, pruning a preset number of parameter groups with the lowest importance, and performing non-uniform pruning on the remaining parameter groups; performing redundancy matching on the multi-modal large model, using a greedy algorithm to solve the optimal batch scheduling sequence, and performing cache and reuse of embedded representation of redundant nodes between adjacent batches according to the redundancy matching result; and performing efficient parameter adjustment after pruning. The application provides a feasible solution for efficient deployment of multi-modal large models in resource-constrained environments such as edge computing devices.

An elimination method for executing instructions that calculate constant values ​​based on load value prediction is disclosed. One aspect of the method involves predicting the value that a load instruction in the first iteration of a loop will load, and executing multiple instructions that occur after the load instruction in the first iteration of the loop to generate multiple results. Each of the multiple results depends only on the value, one or more constant values, values ​​derived from the value and / or one or more constant values, or any combination thereof. The method also includes generating multiple results for the multiple instructions during each iteration in one or more iterations following the first iteration of the loop without re-executing the multiple instructions. Other methods, apparatuses, and systems are also disclosed.

The invention provides an operator elimination method and system of a neural network, electronic equipment and a storage medium, and relates to the technical field of deep learning. Comprising the following steps: determining a target sub-graph structure with an operator elimination demand in a neural network model; determining an operator in an elimination state in the target sub-graph structure as a target elimination operator; traversing all the sub-graph structures until all target sub-graph structures with operator elimination requirements and corresponding target elimination operators are determined; calculating a buffer memory sum of input and output of the target elimination operator; determining a target cache region of the target elimination operator according to the sum of the internal memories of the buffer regions so as to distribute the target elimination operator to the target cache region, and configuring the same cache address for the outputs of all the target elimination operators; determining the jump offset of the target elimination operator; and compiling the target elimination operator based on the jump offset to complete operator elimination. According to the invention, operator elimination can be realized, and the reasoning time of the neural network can be reduced.