Index unified management and automatic generation method based on internet of things and algorithm model

Abstract

The application provides an index unified management and automatic generation method based on Internet of Things and algorithm model, comprising the following steps: S1: establishing a unified index type system and an index rule library; S2: generating related indexes of production line, model algorithm, algorithm model benchmark configuration; S3: unified standardized calculation real-time index; S4: unified index export realizes consistent data caliber of all platforms; S5: realizing general report configuration and display; the method does not need manual configuration of timing task, backend hard coding and writing of SQL script, significantly reduces development and configuration cost, improves index reusability, can quickly adapt to index demand change of various scenes; based on the unified index library, general report configuration is realized, without special person and special customization development, avoiding logical errors and data deviation caused by manual development, and guaranteeing the accuracy and reliability of the report; the same index is ensured to have the same display result at various places by using completely consistent calculation and statistical caliber, and the data authority is improved.

Description

Technical Field

[0001] This invention relates to the field of energy efficiency optimization and control technology in the steel industry, and in particular to a method for unified management and automatic generation of indicators based on the Internet of Things and algorithm models. Background Technology

[0002] Currently, the steel industry has a long production process, complex technological steps, and large total energy consumption, which places high demands on energy efficiency monitoring, algorithm performance evaluation, and energy-saving optimization and management. The industry currently widely uses energy efficiency optimization and management platforms to centrally manage and visualize energy consumption data and algorithm operation data from steel production sites. The existing technical approach involves deploying IoT data acquisition devices at the project site and using a self-developed IoT platform to collect multi-dimensional production and operation data in real time, such as flue gas volume, fan frequency, flow rate, temperature, and pressure. Various energy efficiency optimization algorithm models are then uniformly configured onto the energy efficiency optimization and management platform to complete model display and operational status monitoring. Based on the real-time collected data, the platform performs data display, analysis, report configuration, and visualization, providing data support for energy efficiency analysis, energy-saving optimization, and operational management in steel enterprises. In practical engineering applications, steel companies and on-site project personnel have a continuous and urgent need for business indicators: First, they need to monitor the production line's operating status, process baselines, and key operational indicators such as algorithm model utilization rate, hit rate, and savings rate in real time; second, they need to automatically or quickly generate various business indicators based on model data collection point configuration, production line start-up / shutdown, or parameter changes; third, they need to share unified indicator data across multiple pages and modules to avoid inconsistencies in results caused by differences in data sources and calculation logic; and fourth, they need indicators to support the flexible configuration, viewing, and reuse of energy consumption reports, providing a reliable basis for energy efficiency management decisions.

[0003] However, existing technologies still have the following drawbacks: Relying on manually generated metrics: This requires manually configuring scheduled tasks, coding, and writing SQL to calculate, analyze, and display metrics, resulting in high development and configuration costs, poor reusability, and difficulty in adapting to changes in requirements. The reports are highly customized, but lack versatility and scalability: the report logic is strongly bound to the production line, algorithm model, and data collection point, making it impossible to quickly adapt to the report requirements of different enterprises, production lines, models, and data collection points. This results in low development efficiency and is prone to logical errors and data deviations, affecting accuracy. Low data credibility: Different pages and modules use different configuration logic, statistical scripts and data sources, resulting in inconsistent results for the same indicator, leading to low credibility. Summary of the Invention

[0004] The purpose of this invention is to address the shortcomings of existing technologies by proposing a unified management and automatic generation method for indicators based on the Internet of Things and algorithm models.

[0005] To achieve the above objectives, the present invention adopts the following technical solution: A unified management and automatic generation system for indicators based on the Internet of Things and algorithm models, the system includes an indicator management module, an association generation module, an indicator rule base, an indicator database, etc. The indicator management module is used to build a unified indicator type system and indicator rule library, and to preprocess real-time data from the production site, etc. The association generation module is used to parse the binding information of configuration objects such as production lines, algorithm models, and benchmark parameters, and generate production line operation status indicators, model algorithm operation indicators, and algorithm model benchmark configuration association indicators based on the indicator rule base and binding information, and complete the real-time calculation of indicators.

[0006] This invention also proposes a method for unified management and automatic generation of indicators based on the Internet of Things and algorithm models, including the following steps: S1: Establish a unified indicator type system and indicator rule library; The indicator management module predefines various categories of indicators in the energy efficiency optimization and control platform and assigns a unique code to each category of indicators. At the same time, it saves the calculation formulas, data sources, statistical periods, and other information of various indicators collected in advance into the indicator rule base of the energy efficiency optimization and control platform to form a unified indicator type system. S2: Relevant indicators for production lines, model algorithms, and baseline configurations of algorithm models; Includes the following sub-steps: S21: Parse the binding information of the configuration object; When a user adds / configures a production line, algorithm model, or baseline parameters of an algorithm model on the energy efficiency optimization and management platform, the association generation module reads the binding information of the currently configured object through the object metadata parsing component. Specifically, the production line binding information includes a list of data collection points, host devices, and operating thresholds; the list of data collection points includes real-time values ​​for each data collection point; the host devices include two states: running and not running. The binding information of the algorithm model includes fixed values ​​of the commissioning status collection points and real-time values ​​of the hit judgment collection points; the commissioning status collection points contain two fixed values, which correspond to the commissioning status of the algorithm model and the non-commissioning status of the algorithm model, respectively. The binding information of the baseline parameters of the algorithm model includes baseline consumption, baseline unit price, correction coefficient, and real-time values ​​of the assessment collection points; S22: Generate metrics based on binding information; The indicators include production line operation status indicators, model algorithm operation indicators, algorithm model baseline configuration related indicators, and custom indicators. Production line operation status indicators: The correlation generation module compares the real-time values ​​of each collection point in the collection point list with the operation threshold to determine the production line operation status; When the real-time value of the data collection point is greater than the operating threshold, the data collection point is determined to be in the production line operating state; otherwise, it is in the production line stopped state. Using the statistical period in the indicator rule library as the calculation interval, the duration of the data collection points corresponding to the production line operating state within the statistical period is accumulated to obtain the production line running time for that statistical period. The total duration of the statistical period is subtracted from the production line running time to obtain the production line not running time. The production line operating state, production line running time, and production line not running time together constitute the production line operating state indicator. Model algorithm operation index: Using the statistical period in the index rule base as the calculation interval, based on the fixed value of the commissioning collection point, the duration of the host device being in operation and the algorithm model being in commissioning state is accumulated to obtain the model commissioning time; The model operational rate is calculated based on the model operational duration and statistical period, whereby the model operational rate = model operational duration / statistical period. The model hit rate is calculated using the following formula: ; Where R is the model hit rate, For the determination value, when hour, =1; conversely, =0; n is the number of real-time values ​​of the hit judgment collection points within the statistical period, and i is the sequence number of the real-time values ​​of the hit judgment collection points within the statistical period. The real-time value and benchmark value of the data collection point are preset by the indicator rule library to determine the hit rate, and B is the target deviation value of the indicator hit rate configured by the system in advance. The total number of times the value is 0 represents the number of times the model missed. The above-mentioned model deployment time, model deployment rate, model hit rate, and number of model misses together constitute the model algorithm operation indicators; Algorithm model baseline configuration associated indicators: The calculation interval is based on the statistical period set in the indicator rule library. The associated generation module retrieves the consumption before and after the statistical period of the assessment collection points in the energy efficiency optimization and management platform. The pre-statistical period consumption is the actual cumulative consumption of the same assessment collection point within the statistical period before the algorithm model is put into operation; the post-statistical period consumption is the actual cumulative consumption of the same assessment collection point within the statistical period after the algorithm model is put into operation. ; ; Savings rate = Savings amount / Baseline consumption amount; The real-time values ​​of the assessment collection points within the statistical period are compared with the preset over-limit threshold point by point. If the values ​​are greater than the preset over-limit threshold, it is considered an over-limit state; otherwise, it is considered a normal state. The duration of the over-limit state is accumulated to obtain the over-limit duration within the statistical period. The total number of times the over-limit state occurs is counted to obtain the over-limit count. The aforementioned savings amount, savings amount, savings rate, overdue time, and overdue number together constitute the baseline configuration related indicators of the algorithm model; Custom metrics: Users can manually add custom metrics in the metric management module according to their actual needs. These custom metrics include mean, difference, frequency of change, custom formulas, etc.

[0007] S3: Standardized and unified calculation of real-time metrics; The system receives real-time production site data uploaded by IoT data acquisition devices, including real-time values ​​of acquisition points, real-time values ​​of acquisition points for hit judgment, real-time values ​​of acquisition points for assessment, real-time values ​​of acquisition points for commissioning status, real-time energy consumption, temperature, pressure, etc. The indicator management module performs preprocessing on the received real-time production data, such as removing outliers and deduplication. After preprocessing, production line operation status indicators, model algorithm operation indicators, algorithm model baseline configuration related indicators, and custom indicators are calculated in real time based on the method in step S2.

[0008] S4: Unified indicator output ensures consistent data standards across the entire platform; The indicator management module saves the real-time indicators calculated in step S3 to the indicator database. All pages, functional modules, and reports on the entire platform read indicator data from this unified database and use a unified statistical standard, ensuring that the results for the same indicator are completely consistent.

[0009] S5: Enables standardized report configuration and display; Based on a unified indicator library, users can quickly configure and display various data reports according to the generated indicators.

[0010] Compared with the prior art, the beneficial effects of the present invention are as follows: This method establishes a unified indicator type system and indicator rule library, calculates indicators through preset rules, and eliminates the need for manual configuration of scheduled tasks, backend hard coding, and writing SQL scripts. This significantly reduces development and configuration costs, improves indicator reusability, and can quickly adapt to changes in indicator requirements in various scenarios. Based on a unified indicator library, generalized report configuration is achieved, eliminating the need for dedicated personnel for customized development. It can quickly adapt to the report needs of different steel companies, different production lines, different algorithm models, and different data collection points, while avoiding logical errors and data deviations caused by manual development, ensuring the accuracy and reliability of reports. A unified indicator output is set up, and all pages, functional modules, and reports across the platform obtain data from the same indicator library, using completely consistent calculation and statistical methods. This ensures that the same indicator displays the same results everywhere, enhances data authority and user trust, and provides reliable support for management decisions. Attached Figure Description

[0011] Figure 1 This is a flowchart illustrating the steps of the unified management and automatic generation method for indicators based on the Internet of Things and algorithm models of the present invention. Detailed Implementation

[0012] To provide a further understanding of the purpose, structure, features, and functions of the present invention, detailed descriptions are provided below with reference to specific embodiments.

[0013] A unified management and automatic generation system for indicators based on the Internet of Things and algorithm models, the system includes an indicator management module, an association generation module, an indicator rule base, an indicator database, etc. The indicator management module is used to build a unified indicator type system and indicator rule library, and to preprocess real-time data from the production site, etc. The association generation module is used to parse the binding information of configuration objects such as production lines, algorithm models, and benchmark parameters, and generate production line operation status indicators, model algorithm operation indicators, and algorithm model benchmark configuration association indicators based on the indicator rule base and binding information, and complete the real-time calculation of indicators.

[0014] like Figure 1 As shown, this invention also proposes a method for unified management and automatic generation of indicators based on the Internet of Things and algorithm models, including the following steps: S1: Establish a unified indicator type system and indicator rule library; The indicator management module predefines various categories of indicators in the energy efficiency optimization and control platform and assigns a unique code to each category of indicators. At the same time, it saves the calculation formulas, data sources, statistical periods, and other information of various indicators collected in advance into the indicator rule base of the energy efficiency optimization and control platform to form a unified indicator type system. The metrics include mean, model deployment rate, model deployment duration, difference, hit rate, number of model misses, production line runtime, frequency of changes, number of times exceeding limits, duration of exceeding limits, built-in benchmark, built-in formula, custom formula, amount saved, amount saved, and rate of savings.

[0015] S2: Relevant indicators for production lines, model algorithms, and baseline configurations of algorithm models; Includes the following sub-steps: S21: Parse the binding information of the configuration object; When a user adds / configures a production line, algorithm model, or baseline parameters of an algorithm model on the energy efficiency optimization and management platform, the association generation module reads the binding information of the currently configured object through the object metadata parsing component. Specifically, the production line binding information includes a list of data collection points, host devices, and operating thresholds; the list of data collection points includes real-time values ​​for each data collection point; the host devices include two states: running and not running. The binding information of the algorithm model includes fixed values ​​of the commissioning status collection points and real-time values ​​of the hit judgment collection points; the commissioning status collection points contain two fixed values, such as 0 and 1, where 1 represents the commissioning status and 0 represents the non-commissioning status, corresponding to the commissioning status and non-commissioning status of the algorithm model, respectively. The binding information of the baseline parameters of the algorithm model includes baseline consumption, baseline unit price, correction coefficient, and real-time values ​​of the assessment collection points; S22: Generate metrics based on binding information; The indicators include production line operation status indicators, model algorithm operation indicators, algorithm model baseline configuration related indicators, and custom indicators. Production line operation status indicators: The correlation generation module compares the real-time values ​​of each collection point in the collection point list with the operation threshold to determine the production line operation status; When the real-time value of the data collection point is greater than the operating threshold, the data collection point is determined to be in the production line operating state; otherwise, it is in the production line stopped state. Using the statistical period in the indicator rule library as the calculation interval, the duration of the data collection points corresponding to the production line operating state within the statistical period is accumulated to obtain the production line running time for that statistical period. The total duration of the statistical period is subtracted from the production line running time to obtain the production line not running time. The production line operating state, production line running time, and production line not running time together constitute the production line operating state indicator. Model algorithm operation index: Using the statistical period in the index rule base as the calculation interval, based on the fixed value of the commissioning collection point, the duration of the host device being in operation and the algorithm model being in commissioning state is accumulated to obtain the model commissioning time; The model operational rate is calculated based on the model operational duration and statistical period, whereby the model operational rate = model operational duration / statistical period. The model hit rate is calculated using the following formula: ; Where R is the model hit rate, For the determination value, when hour, =1; conversely, =0; n is the number of real-time values ​​of the hit judgment collection points within the statistical period, and i is the sequence number of the real-time values ​​of the hit judgment collection points within the statistical period. The real-time value and benchmark value of the data collection point are preset by the indicator rule library to determine the hit rate, and B is the indicator hit rate target deviation value pre-configured by the system. The total number of times the value is 0 represents the number of times the model missed. The above-mentioned model deployment time, model deployment rate, model hit rate, and number of model misses together constitute the model algorithm operation indicators; Algorithm model baseline configuration associated indicators: The calculation interval is based on the statistical period set in the indicator rule library. The associated generation module retrieves the consumption before and after the statistical period of the assessment collection points in the energy efficiency optimization and management platform. The pre-statistical period consumption is the actual cumulative consumption of the same assessment collection point within the statistical period before the algorithm model is put into operation; the post-statistical period consumption is the actual cumulative consumption of the same assessment collection point within the statistical period after the algorithm model is put into operation. ; ; Savings rate = Savings amount / Baseline consumption amount; The real-time values ​​of the assessment collection points within the statistical period are compared with the preset over-limit threshold point by point. If the values ​​are greater than the preset over-limit threshold, it is considered an over-limit state; otherwise, it is considered a normal state. The duration of the over-limit state is accumulated to obtain the over-limit duration within the statistical period. The total number of times the over-limit state occurs is counted to obtain the over-limit count. The aforementioned savings amount, savings amount, savings rate, overdue time, and overdue number together constitute the baseline configuration related indicators of the algorithm model; Custom metrics: Users can manually add custom metrics in the metric management module according to their actual needs. These custom metrics include mean, difference, frequency of change, custom formulas, etc.

[0016] S3: Standardized and unified calculation of real-time metrics; The system receives real-time production site data uploaded by IoT data acquisition devices, including real-time values ​​of acquisition points, real-time values ​​of acquisition points for hit judgment, real-time values ​​of acquisition points for assessment, real-time values ​​of acquisition points for commissioning status, real-time energy consumption, temperature, pressure, etc. The indicator management module performs preprocessing on the received real-time production data, such as removing outliers and deduplication. After preprocessing, production line operation status indicators, model algorithm operation indicators, algorithm model baseline configuration related indicators, and custom indicators are calculated in real time based on the method in step S2.

[0017] S4: Unified indicator output ensures consistent data standards across the entire platform; The indicator management module saves the real-time indicators calculated in step S3 to the indicator database. All pages, functional modules, and reports on the entire platform read indicator data from this unified database and use a unified statistical standard, ensuring that the results for the same indicator are completely consistent.

[0018] S5: Enables standardized report configuration and display; Based on a unified indicator library, users can quickly configure and display various data reports according to the generated indicators.

[0019] This method eliminates the need for developers to perform customized development and coding, significantly improving report generation efficiency and configuration flexibility, and meeting the needs of report display and reuse in different scenarios. Through the above technical approach, this invention fundamentally overcomes the shortcomings of existing technologies, such as reliance on manual configuration, coding implementation, data inconsistency, poor real-time performance, and low efficiency, and realizes the automatic generation, unified calculation, centralized management, and full-domain reuse of energy efficiency management platform indicators.

[0020] When adding or adjusting production lines, algorithm models, or baseline configurations, the system automatically parses the binding information and generates corresponding related indicators, eliminating the need for manual reconfiguration, coding, and debugging. This reduces manual workload, improves response speed, and meets the needs of steel companies for efficient and intelligent indicator management.

[0021] The present invention has been described in the above-described embodiments; however, these embodiments are merely examples for implementing the present invention. It must be noted that the disclosed embodiments do not limit the scope of the present invention. Conversely, any modifications and refinements made without departing from the spirit and scope of the present invention are within the scope of patent protection of the present invention.

Claims

1. A method for unified management and automatic generation of indicators based on the Internet of Things and algorithm models, characterized in that: Includes the following steps: S1: Establish a unified indicator type system and indicator rule library; The indicator management module predefines various categories of indicators in the energy efficiency optimization and control platform and assigns a unique code to each category of indicators; At the same time, the calculation formulas, data sources, and statistical cycles of various indicators collected in advance will be saved into the indicator rule library of the energy efficiency optimization and management platform to form a unified indicator type system. S2: Relevant indicators for production lines, model algorithms, and baseline configurations of algorithm models; S21: Parse the binding information of the configuration object; S22: Generate metrics based on binding information; The indicators include production line operation status indicators, model algorithm operation indicators, algorithm model baseline configuration related indicators, and custom indicators. S3: Standardized and unified calculation of real-time metrics; S4: Unified indicator output ensures consistent data standards across the entire platform; S5: Enables standardized report configuration and display; Based on a unified indicator library, users can configure and display various data reports according to the generated indicators.

2. The method for unified management and automatic generation of indicators based on the Internet of Things and algorithm models as described in claim 1, characterized in that: In step S1, the indicators include model deployment rate, model deployment duration, hit rate, number of times the model missed, production line running time, frequency of changes, number of times exceeding limits, duration of exceeding limits, built-in formula, custom formula, amount saved, amount saved, and saving rate.

3. The method for unified management and automatic generation of indicators based on the Internet of Things and algorithm models as described in claim 1, characterized in that: The specific details of step S21 are as follows: When a user adds / configures a production line, algorithm model, or baseline parameters of an algorithm model on the energy efficiency optimization and management platform, the association generation module reads the binding information of the currently configured object through the object metadata parsing component. The production line binding information includes a list of data collection points, host devices, and operating thresholds; the list of data collection points includes real-time values ​​for each data collection point; the host devices include two states: running and not running. The binding information of the algorithm model includes fixed values ​​of the commissioning status collection points and real-time values ​​of the hit judgment collection points; the commissioning status collection points contain two fixed values, which correspond to the commissioning status of the algorithm model and the non-commissioning status of the algorithm model, respectively. The binding information of the baseline parameters of the algorithm model includes baseline consumption, baseline unit price, correction coefficient, and real-time values ​​of the assessment collection points.

4. The method for unified management and automatic generation of indicators based on the Internet of Things and algorithm models as described in claim 1, characterized in that: The specific details of step S22 are as follows: Production line operation status indicators: The correlation generation module compares the real-time values ​​of each collection point in the collection point list with the operation threshold to determine the production line operation status; When the real-time value of the data collection point is greater than the operating threshold, the data collection point is determined to be in the production line running state; otherwise, the production line is in the shutdown state. Using the statistical period in the indicator rule base as the calculation interval, the duration of the collection points corresponding to the operating status of all production lines within the statistical period is accumulated to obtain the production line running time of the statistical period; the total duration of the statistical period is subtracted from the production line running time to obtain the production line not running time. The production line operating status, production line running time, and production line not running time together constitute the production line operating status index. Model algorithm operation index: Using the statistical period in the index rule base as the calculation interval, based on the fixed value of the commissioning collection point, the duration of the host device being in operation and the algorithm model being in commissioning state is accumulated to obtain the model commissioning time; The model operational rate is calculated based on the model operational duration and statistical period, whereby the model operational rate = model operational duration / statistical period. The model hit rate is calculated using the following formula: ； Where R is the model hit rate, For the determination value, when hour, =1; conversely, =0; n is the number of real-time values ​​of the hit judgment collection points within the statistical period, and i is the sequence number of the real-time values ​​of the hit judgment collection points within the statistical period. The real-time value and benchmark value of the data collection point are preset by the indicator rule library to determine the hit rate, and B is the indicator hit rate target deviation value pre-configured by the system. The total number of times the value is 0 represents the number of times the model missed. The above-mentioned model deployment time, model deployment rate, model hit rate, and number of model misses together constitute the model algorithm operation indicators; Algorithm model baseline configuration associated indicators: The calculation interval is based on the statistical period set in the indicator rule library. The associated generation module retrieves the consumption before and after the statistical period of the assessment collection points in the energy efficiency optimization and management platform. The pre-statistical period consumption is the actual cumulative consumption of the same assessment collection point within the statistical period before the algorithm model is put into operation; the post-statistical period consumption is the actual cumulative consumption of the same assessment collection point within the statistical period after the algorithm model is put into operation. ； ； Savings rate = Savings amount / Baseline consumption amount; The real-time values ​​of the assessment collection points within the statistical period are compared with the preset over-limit threshold point by point. If the values ​​are greater than the preset over-limit threshold, it is considered an over-limit state; otherwise, it is considered a normal state. The duration of the over-limit state is accumulated to obtain the over-limit duration within the statistical period. The total number of times the over-limit state occurs is counted to obtain the over-limit count. The aforementioned savings amount, savings amount, savings rate, overdue time, and overdue number together constitute the baseline configuration related indicators of the algorithm model; Custom metrics: Users can manually add custom metrics in the metrics management module according to their actual needs.

5. The method for unified management and automatic generation of indicators based on the Internet of Things and algorithm models as described in claim 1, characterized in that: In step S3, the system receives real-time production site data uploaded by the IoT data acquisition device. The indicator management module preprocesses the received real-time production data; After preprocessing, production line operation status indicators, model algorithm operation indicators, algorithm model baseline configuration related indicators, and custom indicators are calculated in real time based on the method in step S2.

6. The method for unified management and automatic generation of indicators based on the Internet of Things and algorithm models as described in claim 1, characterized in that: In step S4, the indicator management module saves the real-time indicators calculated in step S3 to the indicator database. All pages, functional modules and reports on the entire platform read indicator data from this unified database and use a unified statistical standard, so that the results of the same indicator are completely consistent.

7. A unified management and automatic generation system for indicators based on the Internet of Things and algorithm models for implementing the method of any one of claims 1-6, characterized in that: The system includes an indicator management module, an association generation module, an indicator rule base, and an indicator database; The indicator management module is used to build a unified indicator type system and indicator rule library, and to preprocess real-time data from the production site. The association generation module is used to parse the binding information of configuration objects of production line, algorithm model, and benchmark parameters, and generate production line operation status indicators, model algorithm operation indicators, and algorithm model benchmark configuration association indicators based on the indicator rule base and binding information, and complete the real-time calculation of indicators.

Images