33 results about "Heuristics" patented technology

A system and method for dynamically generating alarm thresholds for performance metrics, and for applying those thresholds to generate alarms is described. Statistical methods are used to generate one or more thresholds for metrics that may not fit a Gaussian or normal distribution, or that may exhibit cyclic behavior or persistent shifts in the values of the metrics. The statistical methods used to generate the thresholds may include statistical process control (SPC) methods, normalization methods, and heuristics.

Disclosed is a system and method for classifying units of control and concomitant model units from a larger physical system model. The system enables model-based simulation of decoupled system units for improved simulation efficiency. Further, the system allows for automated reasoning about the classification of said units, producing both machine-interpretable and human-readable, natural language descriptions of the decisions and reasons for said decisions made during the classification process, for use in heuristics and user feedback.

The current invention uses structural data mining methods and systems, combined with partitioning hints and heuristics, to locate high level library functional blocks in a gate level netlist of an integrated circuit (IC). In one embodiment of the invention, the library is created by synthesizing various design blocks and constraints. The method supports characterization matching between a netlist and a library, between libraries and between netlists. The data mining method described herein uses a subgraph growing method to progressively characterize the graph representation of the netlist of the IC. In one embodiment of the invention, alternative hashing is used to perform subgraph characterization. Further, the located high level functional blocks may be used to substitute the corresponding portions of the target netlist having the matched characterizations, and may be annotated accordingly in the resulting netlist.

The invention relates to a TD-SCDMA frequency assignment method. Problems in the TD-SCDMA frequency are solved with the hyper-heuristics genetic algorithm. The problems that an existing method is precocious in convergence and long in operation time are solved. The execution process of the hyper-heuristics genetic algorithm is shown in the figure. According to the hyper-heuristics genetic algorithm, an appropriate strategy profile is selected with the high-layer genetic algorithm to continuously optimize annoyance values of a current frequency assignment scheme, and the network quality is continuously improved. According to the TD-SCDMA interference calculation formula, six optimizing strategies are designed and comprise the greedy strategy, the random strategy and the high-speed searching strategy, the random strategy is prevented from being caught into local optimum, and the searching speed can be increased through the high-speed searching strategy. By means of the TD-SCDMA frequency assignment method, the frequency assignment scheme with small interference can be found within a short time.

Embodiments of the present disclosure provide methods, systems, apparatuses, and computer program products for selecting a test suite for an API. In one embodiment, a computing entity or apparatus is configured to receive test patterns and heuristics, receive an input API, the input API comprising API specifications, parse the input API to extract the API specifications, and based at least in part on the extracted API specifications and the test patterns and heuristics, select a test suite, wherein the test suite is programmatically generated using a machine learning model and comprises one or more test routines, one or more data values, and one or more expected results.

Embodiments of the invention provide systems and methods for fair share allocation of inventory levels throughout a supply chain. According to one embodiment, a first round main Linear Programming (LP) solve can generate an initial solution. Post-processing heuristics for fair sharing can be applied to the first round solve of the main LP. Circular sourcing heuristics can be applied to the first round solve when adjusting for fair sharing allocation requirements. For example, applying the circular sourcing heuristics to the first round solve of the main LP can comprise determining a firmed supply surplus and shortage based on a demand picture from the first round solve of the main LP adjusted for fair sharing. A second round main LP solve can be executed using the fixed inter-organizational transfer variables and fixed supply towards independent demand variables from the post-processing heuristics.

This disclosure relates to modeling an agile team structure such that it aligns with agile principles, achieve synergy and deliver intended business benefits. Current approach to modeling depends on unproven manually arrived patterns that do not predict benefits, are based on limited number of experts utilizing heuristics from personal experience. Once a working model is derived, it is refined over time which is a slow process with no verification of its effectiveness. In accordance with the present disclosure, scalable and customizable location independent agile delivery models can be generated using a palette based user interface such that constraints are optimized. A pre-configured meta-model is chosen and the location independent model is generated given the constraints. A compliance indicator provides a degree of compliance with the agile principles. The model is then evaluated using machine learning models that have been trained by leveraging a knowledge base of successfully implemented agile models.

A method using machine learned, scenario based control heuristics including: providing a simulation model for predicting a system state vector of the dynamical system in time based on a current scenario parameter vector and a control vector; using a Model Predictive Control, MPC, algorithm to provide the control vector during a simulation of the dynamical system using the simulation model for different scenario parameter vectors and initial system state vectors; calculating a scenario parameter vector and initial system state vector a resulting optimal control value by the MPC algorithm; generating machine learned control heuristics approximating the relationship between the corresponding scenario parameter vector and the initial system state vector for the resulting optimal control value using a machine learning algorithm; and using the generated machine learned control heuristics to control the complex dynamical system modelled by the simulation model.

Certain example embodiments relate to techniques for generating reassemblable disassemblies of binaries using declarative logic. A declarative logic programming language (e.g., Datalog) is used to compile reverse engineering, binary analysis, and disassembly rules into a format applicable to an executable program, yielding disassembly of that program. Datalog, for example, can be used as a query language for deductive databases, to facilitate this approach. Certain example embodiments thus involve (1) preparation of an executable for Datalog analysis, (2) inference rules and the application of Datalog for program analysis, including the application of Datalog to the domain of binary reverse engineering and analysis, and (3) the collection of assembly code from the results of the Datalog analysis. These rules can include both “hard rules” and “soft rules” or heuristics, even though standard Datalog does not support the latter.

A method of monitoring operation in a material handling environment is described. The method may include obtaining a data stream representing a 3D scan of the target region based on an output from the LiDAR-based sensor. Further, the method may include obtaining operation specification data. The operation specification data may include data related to a standard operation protocol (SOP) and a predefined heuristic. The predefined heuristic may be associated with an operation to be performed by at least one of a machine and an operator in the material handling environment. Further, the method may include generating a machine learning model. The data stream and the operation specification data may be provided as inputs to train the machine learning model. Further, the method may include determining a performance state associated with an efficiency of execution of the operation by using the machine learning model.

Systems and methods can suggest applications for use by a user of a computing device. The suggestions can be provided on a user interface for the user to select. A suggestion engine can use historical user interactions and contextual data to derive features for a machine learning mode. The machine learning model can determine which application to suggest according to the current context. Multiple models, such as a pattern model and a heuristics model, may be user to generate features for the machine learning model based in user interactions.

The invention discloses a pilot assignment method and device based on an improved intelligent water drop algorithm, and relates to the field of ship pilot assignment. The method includes the following steps in sequence: determining problem constraints according to a problem description, and establishing a pilot assignment model; based on the improved intelligence The water drop algorithm solves the pilot assignment model and finds out the optimal pilot assignment plan; the invention proposes a selection strategy based on the optimal candidate node subgroup for the shortcomings of the intelligent water drop algorithm, such as insufficient heuristics and easy to fall into local optimum. search efficiency and algorithm accuracy. The improved intelligent water drop algorithm is used to solve the pilot assignment model, which effectively improves the accuracy of the solution results.

The invention discloses a cloud workflow segmented online scheduling optimization method integrating heuristics and intelligent computing, which comprises the following steps: obtaining information required for scheduling optimization; calculating the sorting value and level value of tasks; The heuristic method generates the task scheduling optimization scheme of the first stage; obtains the task scheduling optimization scheme of the second stage based on the genetic algorithm; outputs the scheduling optimization scheme. The present invention adopts a segmented scheduling optimization method that combines heuristics and intelligent computing, and the solution time is equal to that of the heuristic method, while the quality of the solution is similar to that obtained by the intelligent computing method. Therefore, the present invention adapts to real-time online scheduling. Effectively improve the quality of understanding under the premise.

A method to improve, simplify, and facilitate multi-dimensional real time vector search and a provide a heuristics backed insight engine tool and a computer system to improve, simplify, and facilitate multi-dimensional real time vector search and provide a heuristics backed insight engine tool to provide real time, free flow, multi-dimensional search or query function of high volume, structured data to traverse to a point in generalized multi-dimensional Euclidean space without constraints of pre-identifying search dimensions.

The invention discloses a behavioral modeling method of a digital twin manufacturing unit based on a time automaton, investigates the physical entity composition and processing requirements of the manufacturing unit, obtains behavioral characteristics and corresponding time parameters, and aims at specific processing tasks based on heuristics The intelligent algorithm generates the running sequence of manufacturing unit processing equipment including the unique identification number, process number, processing equipment identification number and processing time; constructs the time automaton model of processing equipment, transfer equipment, buffer zone and storage; The core constructs a timed automaton network to generate the behavior model of the manufacturing unit; according to the operation sequence of the processing equipment, it generates the overall behavior sequence of the manufacturing unit including the equipment operation sequence and the WIP status tracking sequence, which is used for the execution analysis of the manufacturing unit twin. The invention effectively and accurately describes the behavior logic of the manufacturing unit, and improves the self-organization, self-adaptation and self-decision-making capabilities of the manufacturing unit.

The disclosure relates generally to a parallel computing framework designed to accelerate intensive kernel operations and address the complex physics in numerical reservoir simulations by effectively using the ‘many-core’ computing platform. Specifically, this Many-core Advanced Parallel Simulation (MAPS) uses heuristics to decide the optimal hardware configurations and optimal software components for a given problem's characteristics to efficiently design a model.

The invention provides a sparse technology-based fast Gauss-Seidel power flow algorithm for a power system and belongs to the analyzing and computing field of the power system. According to the technical scheme of the algorithm, data only containing non-zero elements in a data file are quickly read into an array Y(n, d), so that the reading time of a storage unit and the data file is greatly reduced. Vi and the Qi of a PV node are quickly calculated based on the array Y(n, d), so that the invalid computation of a large amount of zero elements is avoided. Meanwhile, an optimal acceleration factor is quickly determined based on the improved heuristics method, so that the iterative computation speed is greatly improved. The above algorithm is completely applicable to the checking calculation process of all IEEE systems, such as an IEEE-118 system. Compared with the prior art, the time spent in reading the data file is reduced by about 85%, and the time spent on the power flow iteration process is reduced by about 90%. Moreover, the entire computation time is reduced by about 88%. In addition, the more the nodes of a system, the greater the advantages of the algorithm.

The invention relates to a method for event-oriented transmission of process measurements from at least one measuring device of a control system of a process plant to an operator station server (4) of the control system (3) of the process plant, comprising: a) receiving the process measurements by means of an automation device (9) of the control system (3) of the process plant; b) automatically sensitizing or desensitizing the measurement range of the process measurement values by means of an automation device (9) of a control system (3) of the process plant, automatically reducing the measurement range during sensitization and expanding the measurement range during desensitization by the automation device (9) by means of a heuristic method according to the fact that the transmission rate is lower than the minimum transmission rate of the process measurement value from the measurement device to the automation device (9) or exceeds the maximum transmission rate of the process measurement value from the measurement device to the automation device (9); c) transmitting the process measurement value to an operator station server (4) of a control system (3) of the process plant in an event-oriented manner, the event representing a predetermined change in the process measurement value.

The invention puts forward a huristic engineering optimization method based on convergence trajectory control. The method comprises following steps: bringing forward a convergence trajectory equation function set for an optimization method; establishing a convergence prediction calculation method for the optimization method; dynamically adjusting parameters to ensure that actual convergence value is the same as the design convergence value of the optimization method. The huristic engineering optimization method based on convergence trajectory control has following beneficial effects: without parameter setting, high optimization efficiency is obtained and good search quality is achieved; generally speaking, the invention is a novel huristic engineering technology and innovative and of great scientific meaning and value for enrichment and improvement of and huristic engineering optimization method, theory and application.

The invention relates to a non-API object recommendation method based on heuristics and neural networks, and belongs to the technical field of code completion and code recommendation. According to themethod, all members included in a non-API object declaration type are collected according to non-API member accessing sample sample on the right side of an assignment statement in open source software, the members comprise members obtained by inheritance, then inaccessible members are eliminated according to the relationship between the class where the assignment statement is located and the non-API object declaration type, and the remaining accessible members are all taken as candidates and put into an initial candidate list cdtList for the use of subsequent use. Prediction is conducted on the sample in step one based on three kinds of specific heuristic rules. The neural networks are trained by using information to obtain a filter capable of filtering low-reliability prediction results.When a programmer enters '.' in a non-API sample object on the right side of the assignment statement, non-API members which might access are predicted. According to the method, the probability thatthe number of recommended members under the same data set is correct is obviously higher than existing methods and tools.