47 results about "Steady state detection" patented technology

Methods and systems to detect steady-state operations in a process of a process plant include collecting process data. The collected process data is generated from a plurality of process variables of the process. A multivariate statistical model of the operation of the process is generated using the process data. The multivariate statistical model may be generated from a principal component analysis. The model is executed to generate outputs corresponding to the most significant variations in the process. Statistical measures of the outputs are generated and used to determine whether a steady-state or unsteady-state is related to the process.

The invention provides a wearable system and method for cardinal symptom quantitative detection of Parkinson patients. The system comprises a glove and a computer, wherein the glove and the computer are connected through wires or in a wireless mode. The glove further comprises a wrist module and a fingertip module. The wrist module and the fingertip module are worn on the fingertips and the wrist of a patient respectively. A tremor detection module, a tremor amplitude steady state detection module, a muscular rigidity detection module and a slow movement symptom detection module are arranged in the computer and are used for tremor detection, tremor amplitude steady state detection, muscular rigidity detection and slow movement symptom detection of the Parkinson patients and displaying detection results. For the situation that the symptom judgment of different neurologists for the same patient is different, the system has the uniform quantitative detection standard. Detection of three main symptoms is integrated in one system, and the system can provide complete and comprehensive illness state detection.

The invention relates to a tobacco manufacturing stability inspection. It solves the issue of failure to effectively identify the factors affecting the tobacco quality, resulting in the inability of reasonable quantitative and qualitative analysis with simple application, effectively getting rid of all kinds of abnormal status, ensuring the quality of tobacco. Using time sequence scuff, distributional statistical chart and accumulated probability curve, separating the process stability from the overall process, judging the conformity of its distribution and that of the normal distribution, effectively identifying the process variation factors, tracking the time sequence zone, analyzing the variation source, type and the affecting degree to the quality index.

The invention discloses a data correction method under the condition of variable load of a chemical flow system. The method comprises the following steps: firstly, acquiring measuring data of a measuring instrument within a period of time, and applying steady state detection technology for steady state determination according to key measuring bits among the measuring data; secondly, taking mean values of the measuring data in various steady states, performing steady state operating condition interval division on the mean values of the steady state data according to the load, and dividing the mean values into a plurality of different steady state operating condition intervals; and thirdly, performing data correction and remarkable error detection on multi-variable multi-group measuring data in each steady state operating condition interval, and obtaining remarkable error detection results and data correction results of various steady state operating condition intervals. The method has more reliable and more accurate correction results, simultaneously can select other steady state detection technology and data correction technology, and is flexible to apply and simple to realize.

The invention discloses a method for determining a thermal power unit operation optimization target value based on Apache Spark and an improved FP-Growth algorithm. The method includes the following steps of: S1, selecting influence parameters of a power supply coal consumption rate, collecting historical operation data according to the influence parameters, and performing data preprocessing and steady-state detection; S2, screening operation optimization parameters from the influence parameters of the power supply coal consumption rate by using Pearson correlation analysis; S3, improving an FP-Growth algorithm based on a matrix technology; S4, parallelizing the improved FP-Growth algorithm based on Apache Spark; S5, discretizing the data of the operation optimization parameters, mining afrequent pattern from a data set by using the parallel improved FP-Growth algorithm, inversely discretizing a mining result, and sorting out the mining result to obtain the target value of the unit operation optimization parameters under various working conditions. The method for determining the target value of the thermal power unit operation optimization parameters disclosed by the invention isbased on the Apache Spark and the improved FP-Growth algorithm, and has the advantages of short time consumption, low memory usage and high efficiency of mining massive data.

The invention discloses an oil refining process online optimization method and an oil refining process online optimization device based on a case library and steady state detection. The oil refining process online optimization method comprises the steps of: acquiring historical data from each preset time length of a database of a factory, generating historical cases and storing the historical cases into a historical case library; acquiring current production data from the database of the factory, and generating a current production case; and performing fuzzy matching on the current productioncase and the historical cases in the historical case library to obtain the historical case which satisfies a preset condition with fingerprint data of a current production state, sorting the historical cases according to assessment data to obtain a referable historical case set superior to the current production case, so as to optimizing the oil refining process according to the historical case set. The oil refining process online optimization method effectively solves the practical problems and defects of the related art, that is, solves the problems of poor real-time performance and precision of data, large calculation cost and the like, can realize real-time online optimization of the oil refining process, and is simple and easy to implement.

A system and method for identifying and responding to exceptional charge events of series-connected energy storage elements.

The invention proposes an optimization method and system of a fuel gas system based on heat value balance, and relates to the field of fuel gas system control. The method comprises improving the production and consumption balance rate of a fuel gas by means of meter calibration and newly-added meters, and tracking the change of the heat value of the fuel gas; obtaining the optimal fuel heat valuerequirements of respective fuel consumption devices by using the thermal efficiency of a heating furnace as an evaluation index and by using a data mining method; performing pinch point analysis to obtain the best supply and use matching network of the fuel gas, and optimizing and reforming a fuel gas pipe network based on this, and constructing an operational optimization model of the fuel gas system by using the fuel heat value requirements, the fuel consumption and the pressure requirements of respective heating furnace as constraints, and giving a recovery scheme of effective components inthe gas, a fuel replenishment scheme and an fuel optimal distribution scheme; performing steady-state detection on the fuel gas system and executing an optimized operation plan on the basis of ensuring the stable operation of the system to achieve optimized and efficient operation of the fuel gas system and to improve the overall economic efficiency of a refinery.

The invention discloses a hierarchical clustering-based system steady state detection algorithm. The algorithm comprises the following steps: for industrial data in a period of continuous time interval, calculating distances between every two types of data to obtain a matrix, finding the minimum distance values in the matrix, combining the minimum distance values into a new type, deleting the new type, carrying out processing to obtain an updated matrix and carrying out iterative computation to obtain an expression matrix of a hierarchical clustering tree; for the expression matrix, carrying out computation from the last iterative numerical value to the first iterative numerical value in sequence so as to obtain clustering result reasonability values of the computations, carrying out solving and computation on the computed clustering result reasonability values to obtain final threshold values, carrying out clustering computation on the industrial data through the final threshold values so as to obtain a final clustering result sequence, and carrying out joint time sequence judgement to obtain system steady state condition. The algorithm disclosed in the invention is strong in applicability and has the effect of avoiding the sharp calculation amount increase phenomenon caused by data dimensionality during the data processing.

The embodiment of the invention discloses a working condition steady state detection method and a process optimization method, and the detection method comprises the steps: S10, obtaining the sampling measurement data of a key variable in a preset time period before the current moment for each key variable in the operation process of a device; S20, performing first-order filtering on the sampling measurement data based on preset high-pass filtering parameters and low-pass filtering parameters to obtain a high-pass filtering value and a low-pass filtering value of each sampling value; S30, based on the high-pass filtering value and the low-pass filtering value, performing statistics to obtain a point steady-state metering rate of the key variable; based on the low-pass filtering value and the average value of the sampling measurement data, obtaining the trend steady-state metering rate of the key variable through statistics; and S40, based on the point steady-state metering rate and the trend steady-state metering rate of each key variable, determining the stable state of the working condition of the device at the current moment. The detection method is simple and has universality, and the detection reliability is improved.

An air dynamic steady state detection system for movement of a cam phaser of an internal combustion engine includes a cam position sensing device and a control module. The cam position sensing device generates a position signal based on a position of the cam phaser of the engine. The control module receives the position signal and applies first and second filters to the position signal to select either a transient or steady state condition. The control module also calculates an estimated air value based on the selection of the transient or steady state condition.

The invention provides a steady-state discrimination method for operation conditions in a continuous production process, and the method comprises the steps: S1, obtaining principal components and corresponding feature values based on operation parameters in a production process through employing a PCA method; S2, determining a polynomial filtering window based on a continuous steady-state segment in the first principal component; S3, carrying out the steady-state detection of each principal component based on each window through employing a polynomial filtering steady-state discrimination method; S4, giving weights to the steady-state detection results of all components based on the corresponding feature values, determining a window steady-state detection result, and obtaining the steady-state discrimination result of the operation conditions in a production process according to the steady-state detection result of the window. When the adjustment of the production process is more frequent, the number of monitoring variables is large and the fluctuation is larger, the method can effectively avoid the impact on the steady-state discrimination from a single variable error or several variable errors through extracting the comprehensive information of data, and improves the adaptability in the industrial process.

The embodiment of the invention provides a real-time optimization method for a process industrial process. The real-time optimization method comprises the following steps: carrying out data validity detection and steady-state detection on a real-time working condition; searching a historical optimization working condition in a historical working condition library to obtain a matching working condition as a reference optimization working condition; performing deviation compensation on the reference optimization working condition by adopting a deviation compensation algorithm; and after deviation compensation, carrying out validity detection on the optimization result, and transmitting the optimization result to the APC in the form of an external target value for control if the validity detection is passed; adopting work condition matching driving, determining a final optimization result of a decision variable in combination with an optimization decision deviation correction algorithm. Online access operation of the mechanism model is not involved in the online real-time operation process; and the optimization scheme of the working condition matching is feasible through actual production verification in history, so that the problem of model calculation convergence cannot occur in the online operation process, the optimization result is real under the condition that the working condition matching is successful, and the long-term operation of the RTO system is ensured.

The invention discloses a multi-target combustion optimization method based on a data-driven fusion strategy. The method comprises the following steps: resampling massive historical operation data ofthe DCS database; performing steady-state detection on the unit load and the coal quality coefficient in the resampling data set; obtaining all time intervals in which the unit load and the coal quality coefficient are in a steady state, wherein the data of each time interval is a steady-state data set; after judgment, enabling each steady-state data set to form a clean data set, combining all theclean data sets to obtain a data set Z0, performing clustering division on unit loads and coal quality coefficients in the data set Z0 to obtain a plurality of working condition partitions, obtaininga combustion optimization rule base and a combustion optimization model base, and applying a strategy 1 and a strategy 2 in the same time. According to the method, the defects of a single data driving strategy are overcome, the influence of real-time working conditions on optimization is comprehensively considered, and the requirements of a coal-fired power plant for real-time performance and effectiveness of multi-target combustion optimization are practically met.

The embodiment of the invention provides a steady state detection method and system for an aluminum oxide production evaporation process. The method comprises the steps of selecting a plurality of steady-state detection variables from the process variables according to the correlation between the process variables in the evaporation process and the concentration of four-flash discharge material; utilizing the improved K-means algorithm to perform outlier detection on a first sample data set corresponding to each steady state detection variable, and obtaining a second sample data set corresponding to each steady-state detection variable; acquiring the change speed and the change acceleration of each steady-state detection variable at any moment in the time period to be detected according tothe second sample data set corresponding to each steady-state detection variable; acquiring a multivariable steady-state evaluation index of any moment according to the change speed and the change acceleration of each steady-state detection variable at any moment, and judging whether the evaporation process is in a steady state at any moment according to the multivariable steady state evaluationindex. The multi-variable information is integrated, the influence of variable errors is reduced, and steady-state detection in the evaporation process is realized.

The embodiment of the invention provides an operation optimization method based on refining device operation data. The method comprises the steps of determining a data type and a state parameter for constructing a refining device case library; extracting historical operation data of the refining device based on the determined data type according to a preset duration, and performing steady-state detection on the extracted historical operation data in combination with the state parameters; selectively carrying out case extraction according to a steady-state detection result, and screening extracted cases to obtain multiple types of case libraries; and taking the operation data of the refining device based on the determined data type, matching the operation data with the operation data storedin the case library, and outputting an operation optimization suggestion for the refining device according to a matching result. By performing steady-state detection and case extraction on historicaloperation data, the application of historical data is truly realized; and by designing different case library types of multiple targets, the number of the targets can be flexibly increased and decreased, and classified application of historical operation data is realized.

The invention provides an LES dual-mode rapid power grid fundamental current amplitude detection method which comprises the following steps of S1, establishing an LES filter universal model; S2, by designing different actual sampling times N and sampling periods Ts, establishing a high-speed LES filter and a low-speed LES filter; S3, respectively feeding a certain phase of power grid voltage signal into the high-speed LES filter and the low-speed LES filter, and obtaining a high-speed LES detection value and a low-speed LES detection value; S4, obtaining an absolute value of a difference value of the high-speed and low-speed LES detection values; S5, feeding the absolute value into a hysteresis comparator, if the absolute value is greater than a hysteresis upper limit, using the high-speed LES detection value as a final value for output, and if the absolute value is smaller than a hysteresis lower limit, using the low-speed LES detection value as the final value for output. According to the dual-mode rapid power grid fundamental current amplitude detection method provided by the invention, power grid fundamental current amplitude detection time is shorter than a quarter of cycle wave, and meanwhile, effectively inhibits influence of low order harmonics on detection accuracy and promotes steady state detection accuracy.

The invention relates to the technical field of artificial intelligence, in particular to a water body steady state detection method and system in airtightness detection based on artificial intelligence. The method comprises the following steps: obtaining a gray level image of a pool, and obtaining a bubble motion region through continuous multi-frame gray level images; obtaining a water surface region of the bubble motion region, and obtaining a water surface connected domain and a water surface connected domain mass center according to the water surface region; filtering the water surface connected domain in a set filtering direction to obtain a water surface bubble connected domain, and determining the relative position of the optical axis of the camera and the water surface according to the mass center of the water surface connected domain and the center of the water surface bubble connected domain; obtaining a first edge straight line in the water surface connected domain, screening out a second edge straight line according to the relative position, and obtaining a water surface edge region; and collecting continuous multi-frame optical flow information of the water surface edge region to judge the steady state of the water body. According to the method, the real water surface edge is obtained in the image, and the water body steady state can be accurately detected.

The present invention is intended to prevent the flow of a large input current into a DC/DC converter, to which power is fed from a power supply, before a supply voltage delivered from the power supply reaches a rated output voltage. A control unit included in the DC/DC converter includes a steady state detection block that detects a condition in which an input voltage to be applied to the DC/DC converter has been stabilized, and inhibits the power feed from the DC/DC converter until the input voltage delivered from a power supply in a preceding stage is stabilized.

The invention relates to an automobile driving wheel surface output power detection method. According to the method, on a chassis dynamometer of an automobile, constant speed loading control detection is performed on the dynamic performance of the automobile based on a speed point corresponding to the rated power point or maximum torque point of an engine; the detected automobile can be weighed, so that frictional resistance power Pfi of the roller of the wheel of the automobile can be obtained, and bench resistance is known quantity stored in a computer, and therefore, the bench resistance power Pti of the speed point can be obtained; when a gearbox is shifted to a specified gear where a maximum accelerator pedal position is located, constant speed loading control steady state is adopted to detect the loading power of Pdi of a power absorption device at a corresponding speed point, and therefore, the driving wheel surface output power Pwi of the corresponding speed point under a standard environmental state can be expressed by the equation that Pwi=(Pdi+Pfi+Pti)*a, wherein a is a power correction coefficient; and in order to reduce the influence of maximum torque speed errors, a torque ratio can be adopted to replace a power ratio to be compared with a limit value.