QUALITY-SENSITIVE CALCULATION SYSTEM AND METHOD FOR INDUSTRIAL DATA
Patent Information
- Authority / Receiving Office
- TR · TR
- Patent Type
- Applications
- Current Assignee / Owner
- SİSKON ENDÜSTRİYEL OTOMASYON SİSTEMLERİ SANAYİ & TİCARET ANONİM ŞİRKETİ
- Filing Date
- 2026-04-30
- Publication Date
- 2026-06-22
Smart Images
Figure 00000024_0000
Abstract
Description
1 TARIFF QUALITY-SENSITIVE CALCULATION SYSTEM FOR INDUSTRIAL DATA AND THE METHOD Technological Field: The invention relates to industrial data processing, data analytics, and industrial automation techniques. It relates to the field; specifically, to machinery, PLCs, sensors, and equipment located in the production area. The quality status of data obtained from similar industrial data sources 10 processing taking this data into consideration, and calculation results based on this data. the production and quality and / or validity of the calculation results in question a calculation that enables the transfer of information to target platforms. It relates to the system and how this system works. The invention relates solely to the processing of data obtained from industrial data sources. No, it depends on determining the quality status of this data, on the quality levels. whether to be included or excluded from the calculation process and The calculation result has a quality that reflects the quality of the data used. and / or a system for creating it together with validity information 20 It includes. The invention also addresses situations involving incomplete, delayed, erroneous, or unreliable data. the calculation processes should continue without interruption and the calculations The results are marked along with quality information and sent to target platforms 25 It also includes a system and method that enables its transfer. State of the Art: In industrial automation systems, the machine, PLC, 30 located in the production area processing data obtained from sensors and similar equipment and this data Generating various computational results using digital methods is a hallmark of modern digitalization. 2 and with the widespread adoption of Industry 4.0 applications, it has become of great importance. has been successful. Within this scope, the data in question are used for monitoring, analysis, and reporting. It is used in predictive maintenance and decision support processes, and in calculations. The results play a critical role in managing production processes. Current practices are carried out using industrial data. Computational processes are mostly statically defined rules and algorithms. or is carried out through predetermined computational models. This In computing systems, calculation engines process data directly from data sources. It produces results by processing, and the quality of the data used is mostly 10 This is not taken into account in the calculation process. This situation is particularly problematic due to data deficiencies. results produced in cases such as data delay or erroneous data generation This leads to a decrease in its credibility. However, in industrial environments, data obtained from data sources 15 It is not always possible for data to be consistent, complete, and timely. due to losses, communication delays, sensor failures, or system errors The data may arrive at the system incomplete, delayed, or inaccurate. This is possible. In current systems, in such cases, either the calculation processes being stopped or any quality issues based on existing data 20 Results continue to be produced without any evaluation. Both In this case, system performance and the reliability of the results are negatively affected. is affected. On the other hand, 25 of the results produced by current calculation systems to establish quality or validity information regarding its reliability There is no integrated approach. The calculation results produced are superior. are transmitted directly to the systems, and it is necessary to determine under what data conditions these results are obtained. how it is produced or how reliable it is is distinguished by higher-level systems. This is not possible. This situation is particularly problematic for automated decision-making mechanisms. 30 It leads to erroneous assessments and incorrect actions in the systems where it is used. It can open. 3 Within the scope of known techniques, assessments regarding data quality are also included. It mostly remains limited to the data processing stage, this information is calculated. not directly integrated into the process and reflected in the calculation results This is observed. Therefore, information regarding data quality is effectively implemented throughout the system. 5 is not being used in any way and will increase the accuracy of the calculation results. It cannot be evaluated in this way. In industrial communication protocols, particularly within the scope of OPC UA, data Quality assessments are mostly in the form of "Good / Bad / Uncertain". Data quality is expressed using categorical quality codes. In this approach, data quality is defined by specific criteria. They are classified according to thresholds, and relate to the behavior of the data. No statistical or continuous evaluation is being carried out. However, the invention... Within this framework, input data are treated as signals, and these signals are... 15 statistical moments such as variance, standard deviation, skewness, and kurtosis continuous and considering sampling design and systematic deviance behavior A quantitative quality score is generated. This allows data quality to be assessed categorically. It is expressed as a numerical measure of reliability rather than a classification, and the word This issue is further integrated directly into the quality score calculation process. A sensitive and integrated assessment opportunity is provided. 20 The literature search revealed the number US20240338349A1. In patent documents, data obtained from different data sources follow a rule. evaluation through the engine and quality as a result of this evaluation or a system for producing outputs that express validity status 25 In this system, rules are applied depending on the data conditions. It is being run and the results obtained are being used for quality assessment purposes. It is used. In this context, rule enforcement based on data exchanges and The results are expressed with a quantitative evaluative value. However, the relevant document states that quality information is directly integrated into the calculation process. 30 by selecting or eliminating input data and applying that selection to the calculation. by systematically reflecting the output quality score on the result 4 It does not offer a solution for determining this. However, the word the system in question is a physical system depending on the reliability of the calculation result. Determining an actuator command that has an effect on the system or It does not include a control layer in the form of a limitation. Therefore, the relevant The document shows that the calculation outputs are directly integrated into physical control loops. 5 and a control at the actuator level depending on the output quality score. It does not enable the creation of a mechanism. The patent numbered US8856058B1 was encountered in the literature search. In the document, information obtained from different data sources is used to create a rule engine. 10 It is evaluated within and the rules are dynamic depending on the data context. A system for its implementation is described. In the system, data objects can be dynamically addressed, and objects associated with this data can be created. This ensures that the rules are enforced during the process. This structure provides data. a rule enforcement mechanism emerges that can vary depending on the circumstances. 15 It implements a data-driven business rule management system instead of static algorithms. It provides this information. However, the relevant document is based on industrial data sources. The quality of the obtained data is directly integrated into the calculation process. the systematic analysis of the impact of low-quality data on the computational process 20 to be created along with quantitative quality and / or validity information It does not include a solution. The patent numbered CN121352630A was encountered during the literature search. In the document, the rule engine uses data from multiple data sources 25 quality refers to the analysis performed through and the reliability of data sources. or a system for calculating score values is described. In this system, quality assessment is carried out based on data conditions. The results obtained are expressed as a quality score. However, the relevant The document directly integrates these quality scores into the calculation process. 30 By selecting the input data, the calculation process is performed according to this selection. the execution and calculation result depend on the quality of the data used. a solution for producing an output with a defined output quality score It does not provide any information. Furthermore, the relevant document does not address the reliability of the calculation results. a control layer that directs the physical system behavior depending on the system or It does not include a control mechanism at the actuator level. However, the relevant document explains the process of calculating these quality scores. By directly integrating and selecting the input data, the calculation process is carried out in this way. the selection is made according to the data used and the calculation result produced with an output quality score determined based on its quality. It does not offer a solution for this. Furthermore, the relevant document states that the calculation result is 10. actuator to be applied to a physical system depending on its reliability It does not include a control layer for determining the commands, output a control mechanism at the actuator level depending on the quality score It does not offer this. Therefore, the system is integrated into physical control loops. It does not present a reliability-based decision-making structure that can be established. 15 In conclusion, the current technology uses data obtained from industrial data sources. low-quality data where quality status of the data is included in the calculation process the impact of the data is controlled, the reliability of the calculation results produced with quality and / or validity information stating this and incomplete or 20 In cases of erroneous data, calculation processes continue without being stopped. The need for an integrated computational system and methodology continues. Brief Description of the Invention: The invention does not relate to the processing of data obtained from industrial data sources, calculation processes taking into account the quality of the data in question inclusion and quality and / or validity information of the calculation results It relates to a calculation system and method for collaborative production. The computational structure presented in this invention is based on known data processing and rules. It is not a simple combination of engine systems, especially for uncertain data. 6 a new technical effect towards reliable and marked result production under these conditions This reveals that generating a quality score for input data is crucial in this context. the subject is the calculation process by evaluating quality scores with threshold values. Determining the data to be included and the calculation result used. 5. Creating an output quality score that reflects the quality status of the data. is provided. The main purpose of the invention is to correct incomplete data obtained from industrial data sources. calculation in case of delay, error or low reliability to prevent the processes from stopping and to express the reliability of the calculation results 10 The goal is to ensure that it is produced with a quality and / or validity information. The system developed within the scope of the invention; obtains from industrial data sources (A) The data obtained is stored in the artificial intelligence module (11) located within the server (10) and processing by the calculation engine (12) data collection interface (20) 15 generating a quality score for the data obtained, and thresholding these quality scores. data to be included in the calculation process by comparing with the values the determination of the data by the calculation engine (12) using the calculation result and the generated calculation (30) 20 together with the quality and / or validity information of the result into the database. by recording and transferring to target platforms (B) It is structured. The invention addresses low-quality data that is not included in the calculation process. 25 The output quality score of the calculation result is determined by taking this information, and In this way, the calculation result, along with the reliability of the result in question, is expressed. Quality information is also produced. The invention also allows the system to function 30 times in cases of incomplete, delayed or erroneous data. the work should continue without interruption and the calculation results should be... Enabled to transfer quality information, along with markings, to target platforms. 7 It offers a working structure. This allows for the elimination of uncertainties in data conditions. This prevents the system performance from being affected and the calculation results from being misused. Its reliability is made available for evaluation by higher-level systems. Another objective of the invention is to determine the quality status of the input data. 5 by controlling the impact of low-quality data on computational results. The goal is to ensure it is brought under control. Another aim of the invention is to present the calculation results not only as numerical values. not as a whole, but as a quality indicator showing under what data conditions these results were produced. and / or to ensure that it is presented together with validity information. Another aim of the invention is to improve data quality in industrial data processing processes. The aim is to ensure more reliable and traceable results are obtained accordingly. Another purpose of the invention is to collect data from data sources in industrial environments. systematic evaluation of the reliability of the data and this The goal is to ensure that assessments are integrated into the calculation processes. Another purpose of the invention is to correct calculation errors caused by data errors or missing data. The goal is to increase system continuity by preventing interruptions to processes. Explaining the Figures: The invention will be described by reference to the attached figures, so that the invention is defined in 25 Its features will be understood more clearly. However, the purpose of this invention is not this specific one. It is not about limiting it with regulations. On the contrary, the invention's accompanying claims all alternatives and modifications that can be included within the area defined by it and the inclusion of their equivalencies is also intended. The details shown are only It is shown for the purpose of describing the preferred arrangements of the present invention and 30 both the shaping of methods and the rules and concepts of invention 8 to provide the most useful and easily understandable description of its properties It should be understood that they are presented in these drawings; Figure 1 - Appearance of the quality-sensitive computing system that is the subject of the invention. Illustrations that will help understand this invention are shown in the attached image. They are numbered and their names are given below. Explanation of References: 10. Server 11. Artificial intelligence module 12. Calculation engine 20. Data collection interface 30. Database 15 40. Control layer A. Industrial data source B. Target platform C. Actuator Description of the Invention: This detailed description explains how the invention is derived from industrial data sources. A system for calculating data based on the quality of the collected data. and the method is solely for the purpose of better understanding the subject, any 25 It is explained in a way that will not create a limiting effect. In industrial environments, data is used from different data sources. During the calculation processes, the data was incomplete, delayed, or incorrect. In this case, the reliability of the calculation results decreases and some 30 In these situations, a complete system shutdown leads to significant technical problems. This opens the door, especially when data quality is not taken into account in the calculation process. 9 In these structures, low-reliability data is directly included in the calculations, and This situation leads to the production of erroneous results. Within the scope of the invention... This technical problem involves generating a quality score for the input data, and the quality in question... data whose scores will be evaluated and included in the calculation process 5. Determination and calculation result along with quality and / or validity information. It is resolved by creating it. Within the scope of the invention, the industrial data source (A) is located in the production area. localized and physical process measurement, status, or control data It refers to the systems and equipment that produce data; these data sources are 10 PLC, industrial sensors, production machinery, measuring instruments, automation. It can consist of systems and similar data-generating units. Target platform (B) is the higher level where the data in question is processed and the calculation results are used. These systems include monitoring systems, data analytics platforms, and decision-making systems. support systems, reporting infrastructures, production management systems, corporate 15 resource planning systems and similar software and / or system infrastructures It may include. Within the scope of the invention, data obtained from industrial data sources (A); Temperature, pressure, flow, velocity, position, vibration, energy consumption, status information, alarm 20 numerical or similar information regarding physical and / or operational parameters. It can include categorical data types. This data involves a continuous data stream. It can be obtained in this way, or it can be generated as an event-triggered process. Data from different data sources are subject to different sampling frequencies, different data 25 signals in the form of varying lengths and data formats It can occur. The computational process carried out within the scope of the invention uses industrial data. one or more input data obtained from sources (A) according to certain rules a new output is generated by processing it according to relationships and / or mathematical operations. 30 It refers to the production of data. These calculations include averaging, summation, difference calculation, ratio calculation, derivation parameter generation, threshold data processing operations such as monitoring, situation determination, predictive analysis, and similar processes. This may include monitoring production processes, performance evaluation, such as fault detection, alarm generation, optimization, and decision support processes. It can be used in various application areas. As shown in Figure 1, the system obtains data from industrial data source (A). a data collection interface (20) that enables the collection of data, the data in question a server (10) in which it is processed, an artificial located within the server (10) intelligence module (11) and a calculation engine (12), calculation results a database in which the calculation results are stored (30) and at least one 10 It includes the target platform (B). Within the scope of the invention, the data collection interface (20) from the industrial data source (A) It is the structure that enables the data obtained to be entered into the system, different data 15 capable of receiving data from its sources continuously or on an event basis It is structured. Server (10) is the main processing unit of the system and the data collection interface (20) processing of the data received, conducting quality assessment and It enables the execution of calculation processes. Located within the server (10) 20 The field artificial intelligence module (11) and the computing engine (12) work together It performs the basic functions of the system. The artificial intelligence module (11) is designed to generate a quality score for the input data. It is structured, and the quality score in question is calculated as follows: the input data are treated as signals. evaluation and statistical characteristics of these signals and behavior This is determined by analyzing its characteristics. In this context, the introduction... data obtained from industrial data source (A) and a physical quantity a numerical data set representing related measurements, with a specific sampling arrangement It is considered as a signal obtained within it. 30 11 The signal in question does not consist solely of individual data points, but rather sequentially. a data series consisting of multiple measured values This data series is being evaluated and the behavior of the signal is being assessed. analyses are carried out. In this context, the artificial intelligence module (11) analyzes the signal. Statistical moments such as variance, standard deviation, skewness, and kurtosis are 5 in calculation, determining the consistency of the signal sampling interval and sensor It analyzes the systematic deviation behavior that occurs in the data. These obtained parameters are then processed using predefined weighting coefficients. These 10 data points are combined to create a quality score for the input data. Thus, the quality score depends not only on the instantaneous accuracy of the data value, considering the signal change characteristics, sampling pattern and overall behavior This is determined by taking into account the system's signal-based and statistical analysis. It offers a quality assessment approach. The resulting quality score is determined beforehand. It is compared against a defined threshold value, and 15 that fall below the threshold value data is not included in the calculation process or has low reliability. This ensures that low-quality data results in computational outcomes. Its effect is being brought under control. The calculation engine (12) performs 20 calculations on the input data, taking into account the quality status. is performing the calculation process, and the said calculation process As a result, it produces an output data. In this context, the calculation engine (12), Derived data based on defined relationships and rules between input data. It ensures its production. The output produced by the artificial intelligence module (11) and the computing engine (12) It is structured to determine the output quality score of the data, and the word The subject is the output quality score; the quality scores of the input data used in the calculation. the amount of low-quality data not included in the calculation and the data used It is determined by taking temporal compatibility into consideration. In this context, 30 high-quality In calculations performed with the data, the output quality score is improved, while the low 12 If the impact of high-quality data increases, then the score in question... is being reduced. In this way, the system provides not only the calculation result but also the word. The subject is a quality 5 that describes the data conditions under which the result was produced. and / or generates validity information. The calculation result and its corresponding output quality score are entered into the database (30) They are recorded and transferred to the target platforms (B). Thus, the higher systems, 10 It is able to take this information into account. Within the scope of the invention, the system handles situations involving incomplete, delayed, or erroneous data. The work is not being stopped, and the calculation process continues with the available data. This is being done and the results obtained are being marked along with quality information. This 15 This allows the system to operate continuously, regardless of data conditions. Within the scope of the invention, the system's workflow is the input from the industrial data source (A). obtaining the data and the data collection interface of the input data (20) It starts with the transfer from the server (10) to the server (10). 20 inside the server (10) The artificial intelligence module (11) produces a quality score for each input data. The quality score in question is calculated by considering the input data as signals. evaluation and statistical characteristics of these signals and behavior This is determined by analyzing its characteristics; within this scope artificial intelligence module (11) variance, standard deviation, 25 of the input data signal in calculating statistical moments such as skewness and kurtosis of the signal determining the consistency of the sampling interval and the occurrences in the sensor data These parameters were obtained by analyzing systematic deviant behavior. by combining the input data with predefined weighting coefficients It creates a composite quality score. 30 13 The resulting quality score is compared against a predefined threshold value. Input data below the threshold value is not included in the calculation process. It is not being evaluated or is marked as low-reliability data. Quality score Inputs that are above the threshold value and / or marked as low reliability data are transferred to the calculation engine (12) taking into account the quality conditions 5 and the input data is calculated by the calculation engine (12) in accordance with defined rules, relationships and / or mathematical operations It is generated as a result of calculation. After generating the calculation result, the artificial intelligence module (11), 10 the arithmetic mean of the quality scores of the input data used in the calculation to include input data not included in the calculation in the total input data to determine the ratio and the temporal relationship between the data used in the calculation To determine the agreement, the time intervals between the collection dates of the data in question 15 This determines the output quality score. The calculation includes the following: If the percentage of uncollected data increases, the output quality score is reduced. However, this applies if the data used is temporally inconsistent. The score is reduced accordingly. The invention also includes a system based on the output quality score of the calculation result. actuator (C) commands through a control layer (40) This allows for the predetermination of the output quality score. As a result of comparison with the defined threshold values; a low quality score. The command to be applied to the actuator (C) must be a predefined safe value or 25 setting to the last known safe value, alternative at a medium quality score. Generating a backup command using a computational method and high quality In the score, the command that will be applied directly to the actuator (C) as a result of the calculation This ensures that the data is transmitted as is. Thus, the system only transmits the calculation result. not only produce, but depending on the reliability of the result, physical 30 It also provides a control mechanism that directs the system's behavior. 14 Within the scope of the invention, the control layer (40) is generated within the server (10). calculation result and the output quality score of that result by evaluating the control decisions to be applied to the physical system It is structured as a decision-making and guidance unit that determines and directs oversight. layer (40), output quality score obtained together with the calculation result 5 It compares this with predefined threshold values and uses the result of this comparison. depending on the direct transmission of the command to be applied on the actuator (C), limiting it, replacing it with an alternative calculation result, or This ensures that it is redirected to a predefined safe value. In the scope of the audit layer (40), the reliability of the calculation results is taken into consideration. It creates an intermediary layer that controls the behavior of the physical system by taking control of it. especially under low-quality data conditions, the system must operate securely and in a controlled manner. It provides a mechanism that ensures its operation. The actuator (C) mentioned in the invention is 15 depending on the calculation result. It is a control element that exerts an effect on a physical system, such as a CNC machine. axis drive, spindle drive, valve, motor drive, robot tip or flow control actuators used in various industrial applications such as valves (C) It can include this. In this way, the system does not only produce a calculation result. not only that, but depending on the reliability of the result in question, different physical 20 It enables the implementation of control applications in systems. Within the scope of the invention, the output quality score is calculated using the input data employed in the calculation. Quality scores, input data not included in the calculation, and data used. 25 by mathematical correlation, taking temporal compatibility into account can be determined. In this context, the output quality score (Q_out) is determined based on the input data. Average of quality scores (Q̄_in), proportion of data not included in the calculation. (R_excl) and a time difference parameter (Δt_max) representing temporal alignment using predefined coefficients (α, β, γ) and a time constant (τ) With the help of the following example equation, it can be calculated: 30 Q_out = α·Q̄_in − β·R_excl − γ·(1 − exp(−Δt_max / τ)) In this context, the coefficients α, β, and γ are weighted by the system beforehand. They represent their values, according to different application scenarios. It can be modified. This relationship is used to determine the output quality score. This is an example computational model using different mathematical functions and 5 It can also be achieved using combinations. The calculation result and the output quality score are entered into the database together (30) It is being recorded and transmitted to the target platform (B). Incomplete, delayed or In case of incorrect data, the calculation process is not stopped; the current input is 10. The calculation result is generated based on the data, and the quality of that result is... By marking it with this information, the system flow is ensured to continue. In a sample application, three different sensors located on a production line are used in sequence. 15 where temperature, pressure and flow data are obtained as sequential measurement values This is accepted. The data obtained within this scope are as follows: • Temperature data: 85°C • Pressure data: 12 bar • Flow rate: 30 L / min Each data point is treated as a signal by the artificial intelligence module (11). These signals are evaluated and their statistical characteristics are analyzed to determine quality. Scores are being generated. In this context; • Variance, standard deviation, skewness, and kurtosis in the temperature signal 25 the values are within an acceptable range, the sampling interval is consistent and there was no significant systematic bias in the sensor data. The quality score is determined as 0.95. • High variance in pressure signal, irregular sampling interval and time Due to the detection of increasing systematic deviations, the quality score is 30. It is set at 0.50. 16 • The statistical moments in the flow signal are at an acceptable level, the sampling design is stable and the effect of systematic bias is low This has been determined, and the quality score is calculated as 0.92. Assuming the threshold value defined for the system is 0.70, the pressure Since the data falls below the threshold value, it is not included in the calculation process. 5 Temperature and flow data are included in the calculations. In this example, the calculation engine (12) is used as a process efficiency indicator. It uses the following relationship: Efficiency = Flow Rate / Temperature 10 In this case, the calculation is performed as follows: Efficiency = 30 / 85 = 0.3529 Then the output quality score is calculated by the artificial intelligence module (11). is calculated. In this example, the output quality score is calculated according to the following parameters: is determined as follows: • Average data quality used • Excluded data rate: 20 Calculation: • Average data quality used = (0.95 + 0.92) / 2 = 0.935 • Excluded data ratio = 1 / 3 ≈ 0.33 The output quality score is approximately 0.88 when these parameters are evaluated together. It is obtained as follows. 25 As a result, the system produces the following output: • Calculation result: 0.3529 • Output quality score: 0.88 In this way, the target platform (B) receives not only the calculation result, but also 30 of that result. It can also assess the data quality conditions under which it was produced. Pressure data not included in the calculation due to its low quality, 17 while increasing the reliability of the result; the output quality score is lower due to data deficiency. The goal is to ensure that the value remains below the maximum value.
Claims
18 REQUESTS 1. Using input data obtained from industrial data sources (A) the generation of a calculation result and the target of that calculation result It is a method that enables transfer to platforms (B), and its feature is; 5 • Quality of the mentioned input data by an artificial intelligence module (11) generating the score, • the mentioned quality score with a predefined threshold value comparison, • Input data below the threshold value is not included in the calculation process. not being reported or being marked as low-reliability data, • A calculation using input data that takes quality status into account. The calculation result is produced by the engine (12), • the output quality score of the aforementioned calculation result, in the calculation Quality scores of the input data used and 15 that were not included in the calculation determined by considering the ratio of input data to total input data. • Creating a combined output quality score and target based on the calculation result. It is characterized by including the steps of transferring to platforms (B). It is done.
2. It is a method according to Claim 1, and its characteristic is that the quality score is derived from the input data. considered as signals and the variance of these signals, statistical moments such as standard deviation, skewness, and kurtosis It is determined by calculation.
3. A method according to claim 1 or 2, characterized by its quality score, input. Determining the consistency of the sampling interval of the signals in the data. It is created in this way.
4. A method according to any of the above requirements, characterized by its quality 30. the score is the systematic deviation in the signals of the input data. It is formed by determining the behavior. 19 5. A method that meets any of the above requirements, and whose characteristic is quality. the score, the short-term and long-term fluctuation characteristics of the signal This is determined by evaluating trend changes together.
6. A method that meets any of the above requirements, and whose characteristic is quality. the score, statistical moments, sampling interval consistency and systematic predefined weighting coefficients of parameters relating to deviation behavior It is calculated as a composite score by multiplying and combining these values.
7. A method that meets any of the above requirements, and whose characteristic is; output the quality score, the quality scores of the input data used in the calculation It is determined by taking the arithmetic mean.
8. A method that meets any of the above requirements, and whose characteristic is; output 15 The quality score is the total input data that was not included in the calculation. determining the ratio of the data and how this ratio will reduce the output quality score It is created by taking this into consideration.
9. A method that meets any of the above requirements, and whose characteristic is; output 20 Quality score, timestamps of input data used in calculation determining the differences between them and these differences within a predefined time the temporal correspondence value by comparing it with the intervals It is created by calculation.
10. It is a method according to Claim 1, and its characteristic is the output quality of the calculation result. the score must be evaluated by at least one layer of control (40) and the word The subject will be applied to an actuator (C) depending on the output quality score. It is the determination of the command.
11. This is a method according to claim 10, characterized by its ability to predetermine the output quality score. If it is below a defined lower threshold value, the actuator (C) The command to be executed will be based on a predefined safe value or the last known safe value. It is determined as a safe value.
12. A method according to claim 10 or 11, characterized by its output quality score. 5 if it is between the predefined lower and upper threshold values. an alternative calculation method of the command to be applied to the actuator (C) It is determined using...
13. A method according to any of the above requirements, whose characteristic is; output The quality score being above a predefined upper threshold is 10. In this case, the calculation result will be applied directly to the actuator (C). It is designated as a command.
14. It is a method according to claim 13, the characteristic of which is that the mentioned actuator (C) is a CNC machine tool axis drive, spindle driver, valve, motor driver, robot tip 15 or it is a physical control element such as a flow control valve.
15. It is a method according to claim 1, and its characteristic is that the output quality score (Q_out), The average of the quality scores of the input data used in the calculation (Q̄_in), The ratio of input data not included in the calculation to total input data is 20. (R_excl) represents the maximum temporal alignment between the input data. where time difference (Δt_max) is defined by predefined weighting coefficients (α, β, Using γ) and a time constant (τ); Q_out = α·Q̄_in − β·R_excl − γ·(1 − exp(−Δt_max / τ)) It is calculated using the following relationship: 25 16. Using input data obtained from industrial data sources (A) generating a calculation result and targeting that calculation result It is a system that transfers to platforms (B), and its feature is; • to receive the mentioned input data and transmit it to the server (10) at least one structured data collection interface (20), 30 21 • located within the mentioned server (10), for input data to generate a quality score, to set that quality score against a predefined threshold comparison with the value and input data that falls below the threshold value. not including it in the calculation process or considering it as low-reliability data at least one AI module configured for marking (11), 5 • Quality status determined by the mentioned artificial intelligence module (11) rules, relationships and / or mathematical principles defined on the input data the most structured to produce calculation results in line with the processes a small computing engine (12), • The output quality score corresponding to the aforementioned calculation result is 10 in the calculation. quality scores of input data used and data not included in the calculation by considering the ratio of input data to total input data structured artificial intelligence module mentioned (11), • Store the calculation result and the output quality score together. At least one structured database (30), 15 • Calculation result and output quality score together with at least one target. the aforementioned server (10) is configured to transmit to platform (B) It is characterized by...
17. It is a system according to claim 16, and its characteristic is; the quality score of the input data, 20 input data are treated as signals, and these signals are... statistical moments such as variance, standard deviation, skewness, and kurtosis artificial intelligence structured to generate by calculation It includes module (11).
18. A system according to claim 16 or 17, characterized by its quality of input data. its score, the consistency of the sampling interval of the signals from the input data. an artificial intelligence module structured to create by determining (11) is included. Claim 19 is a system based on any claim from 16 to 18, and its characteristic is; The quality score of the input data, the changes in the signals of the input data. 22 by identifying systematic deviant behavior It includes a structured artificial intelligence module (11).
20. Claim is a system based on any claim from 16 to 19, and its characteristic is; The quality score of the input data, the short-term fluctuation of the signal is 5. evaluating long-term trend changes together with its characteristics artificial intelligence module structured to create (11) It includes.
21. Claim is a system based on any claim from 16 to 20, and its characteristic is; 10 quality score of input data, statistical moments, sampling interval parameters relating to consistency and systematic deviant behavior in advance a composite score is obtained by combining the defined weighting coefficients. artificial intelligence module structured to create (11) It includes. 15 Claim 22 is a system based on any claim from 16 to 21, and its characteristic is; The output quality score of the calculation result, the input used in the calculation calculating by taking the arithmetic mean of the quality scores of the data artificial intelligence module structured to create (11) 20 It includes. Claim 23 is a system based on any claim from 16 to 22, and its characteristic is; the output quality score of the calculation result, excluding items from the calculation by determining the ratio of the input data to the total input data and using this ratio as output 25 by taking into account the calculation in a way that will reduce the quality score It includes an artificial intelligence module (11) structured to create. Claim 24 is a system based on any claim from 16 to 23, and its characteristic is; The output quality score of the calculation result, the input 30 used in the calculation by identifying the differences in the timing of data acquisition and this Temporal alignment by comparing differences at predefined time intervals. 23 artificial intelligence structured to create its value by calculating it. It includes an intelligence module (11). Claim 25 is a system based on any claim from 16 to 24, and its characteristic is; Actuator (C) 5 depending on the output quality score of the calculation result. determines the command to be executed on the mentioned server (10) At least one control layer (40) located between actuator (C) It includes.
26. It is a system according to claim 25, and its characteristic is that the mentioned actuator (C) is a CNC 10 machine tool axis drive, spindle driver, valve, motor driver, robot tip or It is a physical control element, such as a flow control valve.