Cost statistics method based on mine energy loss balance

Abstract

The invention relates to the technical field of mine energy loss cost statistics, in particular to a cost statistics method based on mine energy loss balance, which comprises the following steps: deploying multiple types of sensors in mine equipment and areas to collect data, and carrying out anomaly detection, interpolation supplement and classified storage processing; constructing an energy consumption model by integrating equipment power, load and environmental factors, accounting electric power, fuel and total energy cost according to the energy consumption model, and analyzing the proportion; a genetic algorithm, a particle swarm algorithm and an LSTM model are used to optimize equipment parameters and an energy distribution path, and energy demand and cost are predicted; developing a visual interface to display energy consumption and cost data, generating a decision report, and pushing the decision report to a manager; comparing actual and predicted data regularly, collecting equipment, process and market information, and optimizing an energy loss and cost accounting model and algorithm; according to the method, fine energy management is realized, the utilization efficiency is improved, the cost is reduced, scientific decision support is provided for mine energy management, and the system adaptability and sustainability are enhanced.

Description

Technical Field

[0001] The invention relates to the technical field of mine energy loss cost statistics, and in particular to a cost statistics method based on mine energy loss balance. Background Art

[0002] Mine energy management plays a key role in mine production, and is directly related to the production efficiency, cost control and sustainable development of enterprises. With the continuous advancement of industrialization, the demand for energy of mining enterprises has been rising, and the proportion of energy costs in the total cost of mines has also increased. In this context, achieving refined energy management, improving energy utilization efficiency and reducing costs have become important issues that mining enterprises need to solve urgently.

[0003] In the process of mining production, there are many factors that affect energy consumption, among which the problems of equipment operation status, production process and external environment are particularly prominent. Some mining equipment is seriously aged, its operation efficiency is low, and energy consumption remains high, which undoubtedly increases the energy cost burden of enterprises. At the same time, the irrationality of production technology also leads to energy waste. For example, some process flow designs are not optimized enough, making the distribution and utilization of energy in various links not scientific enough. In addition, the mining production environment is complex and changeable, and environmental factors such as temperature, humidity, and altitude have an impact on equipment energy consumption that cannot be ignored. For example, a high temperature environment may make it difficult for equipment to dissipate heat, thereby increasing energy consumption; the air in high altitude areas is thin, and the resistance of equipment operation increases, which also leads to increased energy consumption.

[0004] Traditional mine energy management models are often extensive, with the following main shortcomings: First, there is a lack of accurate monitoring and analysis methods for energy consumption. Most traditional methods can only perform simple total statistics on energy consumption, and cannot deeply analyze the specific energy consumption of each equipment, each process link and different energy types. It is difficult to accurately grasp the key nodes and main influencing factors of energy consumption, and thus cannot provide accurate data support for the formulation of energy-saving and consumption-reducing measures. Second, energy management decisions lack scientificity and timeliness. Due to the lack of comprehensive and accurate data basis, managers often rely solely on experience when making energy management decisions, and it is difficult to make scientific decisions. In addition, the traditional model cannot reflect the real-time changes in energy consumption and costs in a timely manner, resulting in delayed decision-making and inability to respond to sudden problems and market changes in the production process in a timely manner. Third, it is difficult to adapt to the dynamic changes in mine production. In the process of mine production, equipment failures, process adjustments, and market energy price fluctuations occur from time to time. The traditional energy management model lacks flexibility and adaptability, and cannot respond and adjust these changes effectively in a timely manner, which in turn affects the energy utilization efficiency and economic benefits of mining enterprises.

[0005] In view of the many drawbacks of the traditional mine energy management model, there is an urgent need for a more scientific and efficient energy management method to achieve refined management of mine energy, optimize energy utilization, reduce costs, and improve the scientificity and timeliness of decision-making, thereby enhancing the competitiveness of mining enterprises in the market and promoting the sustainable development of the mining industry. Therefore, a cost statistics method based on mine energy loss balance is proposed to address the above problems. Summary of the invention

[0006] The purpose of the present invention is to provide a method for cost statistics based on mine energy loss balance to solve the problems raised in the above background technology.

[0007] To achieve the above object, the present invention provides the following technical solutions:

[0008] A method for cost statistics based on mine energy loss balance includes the following steps:

[0009] Step S1, data collection and transmission: Install smart meters as power monitors at the power equipment in the mine, set fuel flow meters and equip fuel level sensors on fuel consumption equipment, deploy temperature sensors, humidity sensors, pressure sensors in the mine area, and install operating status sensors on the equipment, and transmit the data collected by all sensors to the data processing center through wired and wireless networks;

[0010] Step S2: Data preprocessing and storage: In the data processing center, anomaly detection methods are used to remove abnormal data points based on data statistical laws and equipment operation logic, and interpolation algorithms are used to supplement missing data. Data is stored in relational databases and distributed file systems according to equipment type, production process, energy type, and time dimension.

[0011] Step S3: Energy loss model construction: for each equipment in the mine, the equipment energy consumption model is constructed by comprehensively considering the equipment power characteristics, load influence, and environmental factors; the mine production process is analyzed, and the process energy consumption model is constructed based on the accumulated energy consumption of equipment in each link and the auxiliary energy consumption calculation; the energy loss cost is obtained by subtracting the effective output energy consumption from the total energy input of the mine, and an energy loss decomposition model is established according to the equipment, process link, and energy type;

[0012] Step S4: Cost accounting: Calculate the electricity cost according to the electricity pricing method of the power company and the data of the power monitor, calculate the fuel cost according to the fuel flow meter data and the unit price of fuel, summarize the total energy cost of the mine and analyze the proportion of each energy cost;

[0013] Step S5: Optimization analysis: Genetic algorithms are used to optimize equipment operating parameters, with the lowest unit product energy consumption cost and the highest energy utilization efficiency as the optimization goals; particle swarm optimization algorithms are used to optimize energy distribution paths, with the lowest energy loss as the optimization goal; LSTM time series prediction models are used in combination with production planning information to predict energy demand and cost under different future production plans and market conditions;

[0014] Step S6: Visualization and decision support: Develop a visualization interface to display equipment energy consumption, process energy consumption, and energy cost data; automatically generate energy management decision support reports based on cost accounting and optimization analysis results and push them to mine managers;

[0015] Step S7: Feedback and continuous optimization: Regularly collect actual energy consumption, cost and equipment operation data and compare them with system prediction results, collect information on equipment failures, production process adjustments, and market energy price fluctuations, and optimize and adjust energy loss models, cost accounting models, and optimization algorithms based on the comparison and collection of information.

[0016] As a preferred solution, the anomaly detection method is based on The anomaly detection method based on the principle is , calculate its mean and standard deviation , like or , it is determined as abnormal data and removed; the interpolation algorithm is the cubic spline interpolation algorithm.

[0017] As a preferred solution, the equipment energy consumption model is constructed as follows: Assume that the equipment power characteristic function is , the load influence function is , the environmental impact correction function is , then the equipment energy consumption model in, is the time variable, is the load variable, is the environment variable vector, Indicates that the device is , and The energy consumption under the conditions, the environmental variable vector includes temperature, humidity, pressure, and altitude; the energy consumption model of the process flow is constructed as follows: The energy consumption of the device is , the auxiliary energy consumption is , then the process energy consumption model , represents the total energy consumption of the process, Indicates the number of equipment in the process flow.

[0018] As a preferred solution, the electricity cost is calculated as follows: Assume that the unit price of electricity is The total power consumption is , electricity cost , when the electricity pricing method includes basic electricity charges , peak and valley electricity charges, power factor adjustment electricity charges The peak-valley electricity fee includes the peak-period unit price. , Peak power consumption , Off-peak hours unit price , off-peak hours , time period unit price , power consumption during normal periods , electricity cost ; The fuel cost accounting method is: Assume the unit price of fuel is The total fuel consumption is , fuel costs The total energy cost is calculated as ,in, For other energy costs.

[0019] As a preferred solution, the genetic algorithm is applied in the following way: the equipment operation parameter encoding adopts binary encoding or real number encoding, and the fitness function is constructed with the lowest unit product energy consumption cost and the highest energy utilization efficiency as the optimization goal. ,in, and is the weight coefficient, is the energy consumption cost per unit product, For energy efficiency, select the operation by roulette wheel , single-point crossover operation or multi-point crossover operation, basic bit mutation operation, iteratively searching for the optimal device operating parameter combination according to a predetermined number of iterations, wherein, For individuals The probability of being selected, For population size.

[0020] As a preferred solution, the particle swarm optimization algorithm is applied in the following way: the energy distribution path is encoded as the particle position in the particle swarm optimization algorithm, the energy loss is minimized as the optimization goal, and the particle speed update formula is , the particle position update formula is ,in, and Particles In the Second and The speed of iterations, is the inertia weight, and is the learning factor, and for A random number, For particles In the The individual optimal position of the iteration, For the group The global optimal position of the iteration, and Particles In the Second and The position of the iteration, initialize the position and velocity of the particle swarm, and set the inertia weight , learning factor and .

[0021] As a preferred solution, the LSTM time series prediction model is applied in the following way: historical energy cost data, production data and influencing factor data are used as input variables. Before entering the model, production planning information including planned mining volume Planned beneficiation volume , Planned equipment operating time Perform normalization processing and assume that the LSTM unit is at time The input gate ,in, is the input gate weight matrix, is the hidden layer output at the previous moment, Input for the current moment, is the input gate bias vector, is the sigmoid function, the forget gate ,in, is the forget gate weight matrix, is the bias vector of the forget gate, the output gate ,in, is the output gate weight matrix, is the output gate bias vector, candidate memory unit ,in, is the candidate memory unit weight matrix, is the candidate memory unit bias vector, memory unit ,in, Represents element-wise multiplication, the hidden layer output , by adjusting the LSTM unit weights and biases through learning and training of historical data, and combining production planning information to predict energy demand and cost under different future production plans and market conditions.

[0022] As a preferred solution, the visual interface display forms include bar charts, line charts, and pie charts; the decision support report push methods include email, mobile phone text messages, or system internal message push.

[0023] It can be seen from the technical solution provided by the present invention that the method for cost statistics based on mine energy loss balance provided by the present invention has the following beneficial effects:

[0024] 1. Realize refined energy management:

[0025] Multi-type sensors collect data comprehensively, covering information such as power equipment, fuel consumption equipment, mine area environment and equipment operation status, making mine energy management no longer extensive; whether it is real-time monitoring of power consumption, precise control of fuel use, or consideration of the impact of environmental factors on equipment energy consumption, all provide a full range of perspectives for mine energy management; this helps to accurately locate the key links and main factors of energy consumption, such as determining which equipment is a major energy consumer and which process links have energy-saving potential, thereby providing detailed data support for the formulation of targeted energy-saving measures;

[0026] Classify and store data and build multiple energy models, including equipment energy consumption model, process energy consumption model and energy loss decomposition model, to deeply analyze the details of energy consumption and loss in mining production; cost accounting and analysis based on these models can accurately determine the proportion of each energy cost, so that enterprises can clearly understand the contribution of different energy sources to the total cost, and then optimize energy procurement and use strategies, realize refined energy management from the overall to the local, from the macro to the micro, and effectively improve energy utilization efficiency;

[0027] 2. Optimize energy utilization and reduce costs:

[0028] The optimization of equipment operating parameters by genetic algorithms aims to minimize the energy consumption cost per unit product and maximize the energy utilization efficiency, and can directly act on the equipment operation level; by adjusting the key parameters of the equipment such as the operating frequency, power setting, and operating time, the equipment can be operated in the best state, reducing unnecessary energy waste and reducing the energy consumption cost per unit product; for example, under the premise of not affecting production efficiency, reasonably reducing equipment power or optimizing operating time can significantly reduce electricity consumption, thereby cutting energy costs;

[0029] The particle swarm optimization algorithm optimizes the energy distribution path and is committed to minimizing energy loss. In the complex energy supply network of the mine, the optimal energy distribution method is found to ensure that energy loss is minimized during transmission and distribution. This avoids energy waste caused by unreasonable energy distribution, improves energy transmission efficiency, and indirectly reduces energy costs.

[0030] The LSTM time series prediction model combines production planning information to provide enterprises with energy demand and cost forecasts under different future production plans and market conditions. Enterprises can use this to plan energy procurement and production arrangements in advance to avoid cost increases caused by insufficient or excessive energy supply. At the same time, when energy market prices fluctuate, they can make reasonable procurement decisions, reduce procurement costs, and achieve dynamic optimization of energy costs.

[0031] 3. Improve the scientificity and timeliness of decision-making:

[0032] The visualization interface presents complex data in the form of intuitive charts, such as bar charts showing equipment energy consumption comparisons, line charts showing energy consumption trends, and pie charts analyzing energy cost structures, allowing mine managers to quickly understand energy consumption and cost conditions and gain insight into key information at a glance. This helps managers quickly discover abnormal situations, such as a sudden increase in energy consumption of a certain device or an abnormal change in the proportion of a certain energy cost, so that they can take timely measures.

[0033] The automatically generated energy management decision support report is based on comprehensive cost accounting and in-depth optimization analysis results, providing managers with a systematic and detailed basis for decision-making; the report not only contains a detailed analysis of the current energy management status, but also includes forecasts and optimization suggestions for future trends, so that managers can comprehensively consider multiple factors when making decisions and avoid blind decisions; timely push reports ensure that managers can obtain the latest information at the first time and make scientific and reasonable decisions quickly, such as adjusting production plans, optimizing equipment maintenance strategies, and formulating energy procurement plans, which effectively improves the overall operational management efficiency of the mine;

[0034] 4. Enhance system adaptability and sustainability:

[0035] Regularly collect actual data and compare them with the system's predicted results, as well as collect various relevant information, so that problems and deficiencies in the system can be discovered in a timely manner. Whether it is energy consumption anomalies caused by equipment failure, changes in energy consumption caused by production process adjustments, or the impact of market energy price fluctuations on costs, they can all be captured by the system in a timely manner through this feedback mechanism.

[0036] The energy loss model, cost accounting model and optimization algorithm are optimized and adjusted according to the feedback information, so that the system can continuously adapt to various dynamic changes in the mining production process; this ensures that the system always maintains a high level of accuracy and effectiveness, and can continue to provide reliable support for mine energy management; with the development of mining production and changes in the external environment, the system can continue to evolve, and help mining companies achieve efficient energy utilization, effective cost control and sustainable development goals in the long term, and maintain their advantages in long-term market competition. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] Figure 1 The present invention is a flow chart of a method for cost statistics based on mine energy loss balance. DETAILED DESCRIPTION

[0038] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0039] In order to better understand the above technical solution, the above technical solution will be described in detail below in conjunction with the accompanying drawings and specific implementation methods.

[0040] like Figure 1 As shown, an embodiment of the present invention provides a method for cost statistics based on mine energy loss balance, comprising the following steps:

[0041] Step S1, data collection and transmission: Install smart meters as power monitors at the power equipment in the mine, set fuel flow meters and equip fuel level sensors on fuel consumption equipment, deploy temperature sensors, humidity sensors, pressure sensors in the mine area, and install operating status sensors on the equipment, and transmit the data collected by all sensors to the data processing center through wired and wireless networks; specifically, the following steps are included:

[0042] Step S1-1: Operations related to data collection of power equipment and fuel consumption equipment:

[0043] Power equipment data collection preparation:

[0044] Select appropriate locations to install smart meters on various types of power equipment in the mine to ensure that they can accurately measure the voltage, current, power and other parameters of the power equipment; after installation, debug and calibrate the smart meters to ensure that they work normally and the measurement accuracy meets the requirements;

[0045] Install a fuel flow meter on the fuel supply pipeline of fuel consuming equipment (such as transport vehicles, fuel boilers, etc.), and install a fuel level sensor on the fuel storage container (fuel tank); after installation, commissioning and calibration are also required to ensure that the sensor can accurately measure the fuel flow and level changes;

[0046] Device connection and data transmission settings:

[0047] Connect the smart meter to the mine's internal network via a suitable communication cable (such as an Ethernet cable or a specific industrial communication cable), and configure its data transmission parameters, including the transmission frequency (such as setting it to transmit data every 5 minutes) and the data format (such as using the industrial standard Modbus format), to ensure that the power data can be accurately and timely transmitted to the data processing center;

[0048] The fuel flow meter and liquid level sensor are connected to the data acquisition module, which converts the analog signals collected by the sensors into digital signals. Then, according to the network environment of the mine, a suitable transmission method (such as wired RS-485 bus or wireless GPRS network) is selected to transmit the data to the data processing center. At the same time, the corresponding receiving program and storage path are set in the data processing center to perform preliminary processing and storage on the received power and fuel data.

[0049] Step S1-2: Data collection and transmission of mining area environment and equipment operation status:

[0050] Sensor deployment and installation:

[0051] In the mining area, according to the environmental monitoring needs and regional characteristics, environmental sensors such as temperature sensors, humidity sensors, and pressure sensors are reasonably distributed; for example, they are installed in key locations such as different operating areas, ventilation openings, and equipment rooms in the mining area to ensure that they can accurately reflect the overall environmental conditions of the mine; at the same time, operating status sensors such as vibration sensors, speed sensors, and switch sensors are installed on various types of equipment in the mine to comprehensively monitor the operating status of the equipment. During installation, ensure that the sensors are firm and do not affect the normal operation of the equipment;

[0052] Data transmission and processing preparation:

[0053] Connect environmental sensors and equipment operation status sensors to data acquisition nodes, which can be distributed data collectors or microprocessors integrated in sensors. These nodes are responsible for collecting sensor data and transmitting the data to the data processing center through pre-set wired (such as laying communication cables) or wireless networks (such as Wi-Fi, ZigBee, etc., select the appropriate frequency band according to the actual situation and configure the communication protocol to avoid interference);

[0054] In the data processing center, a classified storage mechanism for environmental data and equipment operation status data should be established. According to information such as sensor type, location, and collection time, the data should be accurately stored in the corresponding database table to provide a reliable data basis for subsequent in-depth analysis and decision-making. At the same time, a data integrity and accuracy check procedure should be set up to ensure that the received data is complete and correct. If any abnormality is found, an alarm should be issued in time to notify relevant personnel to handle it.

[0055] Step S2: Data preprocessing and storage: In the data processing center, anomaly detection methods are used to remove abnormal data points based on data statistical laws and equipment operation logic, and interpolation algorithms are used to supplement missing data. Data is stored in relational databases and distributed file systems according to equipment type, production process, energy type, and time dimension. The specific steps are as follows:

[0056] Step S2-1: Data anomaly detection and processing:

[0057] Anomaly detection method selection and parameter calculation:

[0058] Based on Principle-based anomaly detection method; for each set of data series (such as power data series of power equipment, fuel consumption data series, etc.), first calculate its mean and standard deviation ; To calculate the mean, add all the data values ​​in the data series and divide by the number of data points; to calculate the standard deviation, use the mathematical formula for the standard deviation to first calculate the square of the difference between each data point and the mean, add these squared differences, divide by the number of data points minus 1, and then take the square root of the result;

[0059] Set the abnormal judgment threshold; In principle, if the data point or , then the data point is judged to be abnormal data; the mean here and standard deviation It is calculated for the data sequence currently being processed;

[0060] Abnormal data removal:

[0061] Traverse each data point in the data sequence and compare it with the calculated abnormal judgment threshold; if the data point meets the abnormal condition, that is, xi<μ-3σ or xi>μ+3σ, then remove the abnormal data point from the data sequence; when removing the abnormal data point, record the position of the abnormal data point (such as the timestamp or the serial number in the data sequence) and related information (such as the device type, data type, etc.) to facilitate subsequent analysis of the cause of the abnormality;

[0062] Abnormal data recording and analysis (optional):

[0063] Store the removed abnormal data points and their related information in a special abnormal data record table; conduct a preliminary analysis of the abnormal data to determine the possible causes of the abnormality, such as sensor failure, sudden abnormal operation of the equipment, external interference, etc.; if the number of abnormal data is large or the abnormal situation is serious, it may be necessary to promptly notify the relevant maintenance personnel to inspect and repair the equipment or sensors to ensure the accuracy of subsequent data;

[0064] Step S2-2: Missing data processing:

[0065] Missing data identification:

[0066] In the data sequence, the location of missing data is identified by checking the continuity of the timestamp or serial number of the data points; for example, if the data sequence is arranged in chronological order and it is found that data points are missing in a certain time period, it is determined that there is missing data in this time period; at the same time, the range of the missing data (start time and end time or start serial number and end serial number) and the data type and device to which it belongs are recorded;

[0067] Interpolation algorithm selection and application:

[0068] The cubic spline interpolation algorithm is selected to supplement the missing data. For intervals with missing data, the estimated value of the missing data point is calculated using the cubic spline interpolation formula based on the existing data points before and after the interval. The cubic spline interpolation algorithm constructs a piecewise cubic polynomial function so that the function and its first-order and second-order derivatives are continuous in each piecewise interval, thereby being able to reasonably estimate the missing data while ensuring data smoothness.

[0069] Fill the calculated estimated values ​​of missing data points into the corresponding missing positions in the original data sequence to obtain a complete data sequence; during the filling process, ensure that the time sequence or logical sequence of the data is correct and consistent with other related data;

[0070] Step S2-3: Data classification storage:

[0071] Data classification rule making:

[0072] Formulate data classification rules based on equipment type, production process, energy type and time dimension; for example, equipment type can be divided into mining equipment, transportation equipment, mineral processing equipment, etc.; production process can include mining, crushing, grinding, flotation, etc.; energy type can be divided into electricity, fuel oil, coal, etc.; time dimension can be divided into different time granularities such as day, week, month, year, etc.;

[0073] Relational database storage structure design:

[0074] Design the table structure of the relational database according to the data classification rules; create different tables to store different types of data, such as creating an equipment operation data table, including fields such as equipment ID (associated equipment type), timestamp, and operation parameters (such as power, speed, etc.); create an energy consumption data table, including fields such as energy type, consumption time, and consumption quantity; create a production process data table, including fields such as process link name, time period, and related parameters (such as output, quality indicators, etc.); ensure the relevance and integrity of data by establishing appropriate primary key and foreign key relationships, so as to facilitate data query and analysis;

[0075] Distributed file system storage configuration:

[0076] For some large-scale, unstructured or semi-structured data (such as backup of raw sensor data, detailed logs of equipment operation status, etc.), choose a suitable distributed file system (such as Hadoop distributed file system HDFS) for storage; in the distributed file system, create a corresponding folder structure to correspond to the classification of data; for example, create different folders according to device type, and create subfolders according to the time dimension under each device type folder, and store data files belonging to the device type and time range in the corresponding folder; at the same time, configure the relevant parameters of the distributed file system, such as data block size, number of copies, etc., to ensure the reliability and efficiency of data storage;

[0077] Data storage and indexing:

[0078] The complete data after exception processing and missing data supplementation is stored in the relational database and distributed file system according to the classification rules; in the relational database, the data is inserted into the corresponding table using SQL statements or database management tools; in the distributed file system, the data file is stored in the specified location using the file upload operation; at the same time, in order to improve the efficiency of data query, appropriate indexes are established in the relational database for commonly used query fields (such as device ID, timestamp, etc.) so that the required data can be quickly located and obtained in the subsequent data query and analysis process;

[0079] Step S3: Energy loss model construction: for each mine equipment, the equipment power characteristics, load impact, and environmental factors are comprehensively considered to construct an equipment energy consumption model; the mine production process is analyzed, and the process energy consumption model is constructed based on the cumulative energy consumption of equipment in each link and the auxiliary energy consumption calculation; the energy loss cost is obtained by subtracting the effective output energy consumption from the total energy input of the mine, and an energy loss decomposition model is established according to the equipment, process link, and energy type; the specific operation steps are as follows:

[0080] Step S3-1: Equipment energy consumption model construction:

[0081] Equipment power characteristic function determination:

[0082] For each equipment in the mine, analyze its power characteristics; determine the equipment power characteristic function through monitoring and analysis of technical information provided by the equipment manufacturer, equipment test data or actual operation data. (in, For the device at the time Power is used to indicate the work done per unit time, reflecting the rate at which a device consumes electrical energy or other energy and converts it into other forms of energy; It is a time variable, which is used to describe different moments in the operation of the equipment. It is a dimension of the change of the equipment's operating status. By analyzing the power at different time points, we can understand the dynamic changes of the equipment's power in the entire operation cycle. For a time The function is used to describe the change law of device power over time. Different types of devices have different power change laws over time. The specific form may be a linear function, a polynomial function, an exponential function, etc., which needs to be determined according to the actual characteristics of the device);

[0083] In the process of determining the power characteristic function, the influence of different working modes of the equipment (such as startup, normal operation, standby, etc.) on the power should be considered, and the power characteristic sub-functions under different working modes should be established respectively or the piecewise function should be used to accurately describe the change of equipment power over time;

[0084] Load impact function analysis:

[0085] Study the impact of equipment load on energy consumption and determine the load impact function (in, The degree to which the energy consumption of the equipment is affected by the load, indicating the change ratio or multiple relationship of the energy consumption of the equipment relative to the reference state (such as no-load or rated load, etc.) under different load conditions; Load variables can be parameters related to the load, such as the equipment's processing capacity (such as ore processing capacity, transportation capacity, etc.), work intensity (such as motor load rate), etc. The load variables directly affect the equipment's workload, and thus affect the equipment's energy consumption. The energy conversion efficiency and loss inside the equipment will be different under different load levels; For a load variable The function is used to describe the quantitative relationship between load and energy consumption. For some devices, energy consumption may have a simple linear relationship with load, but for many actual devices, this relationship may be more complex, and may be a quadratic function or other nonlinear relationship. This depends on the physical characteristics and working principle of the device, and the specific function form needs to be determined through experimental testing and data analysis);

[0086] Derivation of environmental impact correction function:

[0087] Consider the impact of environmental factors (temperature, humidity, pressure, altitude, etc.) on equipment energy consumption and establish an environmental impact correction function (in, It is the correction coefficient of environmental factors on equipment energy consumption, which is used to adjust the energy consumption calculation of equipment under different environmental conditions to more accurately reflect the actual energy consumption; It is an environmental variable vector, including environmental parameters such as temperature, humidity, pressure, and altitude. These environmental factors will affect the energy consumption of the equipment through multiple ways, such as affecting the heat dissipation performance of the equipment, air density (which in turn affects the operating resistance of the equipment, etc.), and the insulation performance of the electrical equipment; is a vector of environment variables The function is used to describe the relationship between environmental factors and energy consumption correction coefficients. Different environmental factors have different degrees and ways of affecting energy consumption. Therefore, The specific form of is relatively complex and may be a multivariate function of multiple environmental variables. Its functional expression is determined through a large amount of experimental data and theoretical analysis);

[0088] Equipment energy consumption model integration:

[0089] Integrate the above-determined equipment power characteristic function, load impact function and environmental impact correction function into the equipment energy consumption model ,in, For the device at the time , the load is , environmental conditions are Energy consumption at a certain time, which indicates the total energy consumed by the device under specific operating conditions. It is a comprehensive measure of the energy consumption of the device over a period of time. The integral calculation takes into account the change of power over time and the influence of load and environmental factors;

[0090] Step S3-2: Construction of process energy consumption model:

[0091] Process analysis and equipment energy consumption calculation:

[0092] Analyze the production process of the mine in detail and divide it into several clear process links (such as mining, transportation, crushing, grinding, flotation, dehydration, etc.); for each process link, determine the equipment list involved;

[0093] According to the established equipment energy consumption model, calculate the energy consumption of each equipment in the process flow; for each equipment, obtain relevant data such as its running time, load changes, and environmental conditions in the process flow, and substitute them into the equipment energy consumption model to calculate the energy consumption of the equipment in a specific process link. (in, For the process The energy consumption of a device reflects the energy consumed by the device to perform its specific tasks in the entire process flow. It is an important indicator for evaluating the energy utilization efficiency and cost of each device in the process flow. Indicates the equipment serial number in the process flow. ,in, is the number of equipment in the process. By traversing the serial numbers of different equipment, the energy consumption of all equipment in the entire process can be calculated and summarized);

[0094] Auxiliary energy consumption determination:

[0095] Identify and calculate auxiliary energy consumption in process flows (in, Auxiliary energy consumption in the process, such as lighting electricity, ventilation system energy consumption, equipment preheating energy consumption, etc. Although these auxiliary energy consumptions are not directly involved in the production process of the product, they are essential to maintain the normal operation of the entire process. They need to be taken into account when calculating the total energy consumption in order to comprehensively evaluate the energy consumption of the process);

[0096] Establishment of process energy consumption model:

[0097] Accumulate the energy consumption of equipment in each link and add the auxiliary energy consumption to build the process energy consumption model (in, It is the total energy consumption of the entire process. It is the sum of all energy consumption in the entire process from raw material input to final product output, including the energy consumption of each device and the energy consumption of auxiliary systems. It is a key parameter for measuring the energy efficiency and cost of the process, and provides an overall energy consumption data basis for optimizing the process and reducing energy consumption. Indicates the number of equipment in the process);

[0098] Step S3-3: Energy loss decomposition model construction:

[0099] Energy input and effective output data acquisition:

[0100] Determine the total energy input sources of the mine (such as total electricity supply, total fuel purchases, coal and other energy inputs), and obtain accurate energy input data through corresponding monitoring equipment (such as electric meters, flow meters, etc.); at the same time, determine the effective output energy form in the mine production process (such as the energy content of qualified mineral products, external power output, etc.) and its corresponding quantity or energy value;

[0101] Assume the total power input is (in, The total amount of electricity obtained by the mine from the external power grid, which is monitored and accumulated in real time through the electric meter installed at the power access point, and is one of the basic data for calculating power cost and energy loss);

[0102] The total fuel input is (in, Measured by flow meters installed at fuel storage tanks or refueling points, reflecting the total amount of fuel purchased and used by the mine in a certain period, which is used to calculate fuel costs and evaluate the contribution of fuel to energy consumption);

[0103] Other energy inputs are (in, Used to comprehensively count the energy input of mines, covering all forms of energy used in the mine production process);

[0104] For effective output energy, the energy content of qualified mineral products is (in, It depends on the quality (such as ore grade) and output of mineral products. It is calculated by analyzing the physical and chemical properties of mineral products and output statistics, combined with energy conversion formulas, and represents the useful energy contained in the final product during the mining production process);

[0105] The external power output is (in, If the mine has excess electricity to export (such as electricity generated by power generation equipment and connected to the grid), the output electricity is measured by an electric meter. This part of the output electricity can be regarded as the effective part of the mine's energy utilization and needs to be deducted from the total energy input when calculating energy loss);

[0106] Energy loss cost calculation:

[0107] The energy loss cost is calculated by deducting the effective output energy consumption from the total energy input of the mine; the energy value of energy loss is converted into a cost value according to the market price of energy (such as the unit price of electricity, the unit price of fuel, etc.);

[0108] Assume the unit price of electricity is (in, Determined by the electricity supplier based on market supply and demand, power generation costs and other factors, and used to calculate the contribution of the electricity portion of the electricity cost and energy loss cost);

[0109] Fuel unit price is (in, It changes according to the fluctuation of fuel market price and is a key parameter for calculating fuel cost and evaluating fuel energy loss cost);

[0110] Other energy prices are (in, Used to calculate the cost and loss cost of the corresponding energy);

[0111] The power loss is (in, It represents the net loss of electricity in the mine during the production process, that is, the part of the input electricity that is not converted into effective output (such as product energy or external output electricity));

[0112] The cost of power loss is (in, Obtained by multiplying the unit price of electricity by the amount of electricity loss, reflecting the economic cost caused by electricity loss);

[0113] The fuel consumption is (in, The fuel loss is obtained by deducting the amount of fuel actually effectively used in the production process from the total amount of fuel input; is the effective utilization of fuel);

[0114] The fuel loss cost is (in, The calculation method is similar to the cost of power loss and is used to measure the economic losses caused by fuel loss);

[0115] Other energy loss costs (in, The loss cost of each other energy source is obtained by multiplying the corresponding energy unit price by the loss amount);

[0116] Total energy loss cost (in, Adding up the loss costs of electricity, fuel and other energy sources, we get the total energy loss cost of the mine, which fully reflects the energy utilization efficiency and economic cost of the mine production process);

[0117] Energy loss decomposition model establishment:

[0118] Establish an energy loss decomposition model according to equipment, process flow links, and energy types; for energy loss decomposition at the equipment level, calculate the proportion of each device in energy loss based on the energy consumption model and actual operation data of each device; at the process flow link level, determine the energy loss contribution of each process link based on the process flow energy consumption model; for energy type level, calculate the proportion of different energy sources such as electricity, fuel oil, and coal in total energy loss;

[0119] Equipment The proportion of energy loss is (in, Through computing devices The ratio of energy consumption to total energy loss is used to analyze the relative importance of each device in energy loss and identify the devices with large energy loss so as to take targeted energy-saving measures);

[0120] Process The energy loss contribution is (in, By process The ratio of energy consumption to total energy loss is used to evaluate the proportion of energy loss in different process links, providing a basis for optimizing the process and reducing energy loss);

[0121] Energy Type The proportion of total energy loss is (in, By calculating the energy type The ratio of energy loss to total energy loss is used to analyze the contribution of different energy types to energy loss, which helps to formulate reasonable energy management strategies, such as optimizing energy structure, etc.);

[0122] Step S4: Cost calculation: Calculate the electricity cost according to the electricity price calculation method of the power company and the data of the power monitor, calculate the fuel cost according to the fuel flow meter data and the unit price of fuel, summarize the total energy cost of the mine and analyze the proportion of each energy cost; the specific steps are as follows:

[0123] Step S4-1: Electricity cost calculation:

[0124] Electricity pricing analysis and data acquisition:

[0125] Obtain information from the power company about the electricity pricing method to determine whether it includes pricing items such as basic electricity charges, peak and valley electricity charges, and power factor adjustment electricity charges, as well as the specific pricing rules for each item;

[0126] Obtain power consumption data from power monitors, including power consumption during different time periods (peak, off-peak, and normal periods); ensure the accuracy and completeness of power monitor data, and promptly troubleshoot and repair any missing or abnormal data;

[0127] Electricity cost calculation (single pricing method):

[0128] If the electricity pricing method is only a single unit price, the electricity unit price is ( represents the price per unit of electricity), the total electricity consumption is ( represents the total amount of electricity consumed by the mine in a certain period of time), then the electricity cost ( represents the mine electricity cost calculated according to a single unit price);

[0129] If the electricity pricing method includes basic electricity charges , peak and valley electricity charges, power factor adjustment electricity charges hour:

[0130] Peak-valley electricity fee calculation: Assume the peak period unit price is ( represents the price per unit of electricity during peak hours), the electricity consumption during peak hours is ; The unit price during valley period is ( represents the price of electricity per unit during off-peak hours), the electricity consumption during off-peak hours is ; The unit price for the period is ( represents the price per unit of electricity during normal times), the electricity consumption during normal times is ; then the peak and valley electricity charges ( Indicates the peak and valley electricity fee calculated based on the unit price and electricity consumption during different peak and valley periods);

[0131] Power factor adjustment electricity charges ( It indicates the power factor adjustment fee caused by factors such as power factor. The specific calculation method varies according to the region and power company regulations, and is usually related to factors such as power factor and power consumption. It is determined according to the power factor adjustment coefficient and calculation method provided by the power company;

[0132] Electricity costs ;

[0133] Step S4-2: Fuel cost calculation:

[0134] Fuel data collection and unit price determination:

[0135] Obtain total fuel consumption data through the fuel flow meter ( Indicates the total amount of fuel consumed by the mine in a certain period of time), ensuring that the flow meter measures accurately and the data transmission is normal;

[0136] Determine fuel unit price ( Indicates the price per unit of fuel). The unit price of fuel can be determined based on the market purchase price or the contract price signed with the supplier. Price fluctuation factors must be taken into account. If necessary, the unit price data can be updated regularly.

[0137] Fuel cost calculation:

[0138] Fuel costs ( represents the cost of fuel consumed by the mine);

[0139] Step S4-3: Total energy cost summary and analysis:

[0140] Total energy cost calculation:

[0141] Calculating the total energy cost of a mine (in, Other energy costs, such as coal and natural gas, are calculated in a similar way to electricity and fuel costs, based on the unit price and total consumption of the corresponding energy. represents the total cost of all energy consumed by the mine);

[0142] Analysis of the proportion of each energy cost:

[0143] Calculate the electricity cost ratio ( represents the proportion of electricity cost in total energy cost);

[0144] Calculate fuel cost share ( represents the proportion of fuel cost in total energy cost);

[0145] Calculate the proportion of other energy costs (Calculate the proportion of each other energy separately, Indicates the proportion of other energy costs in total energy costs);

[0146] By analyzing the proportion of each energy cost, we can understand the energy consumption structure of the mine and find out the energy types with higher cost, which can provide a basis for energy management and cost control. For example, if the proportion of electricity cost is too high, we can consider optimizing the operation mode of electrical equipment, adopting energy-saving equipment or exploring the use of renewable energy to reduce electricity costs.

[0147] Step S5: Optimization analysis: Use genetic algorithms to optimize equipment operating parameters, with the lowest unit product energy consumption cost and the highest energy utilization efficiency as the optimization goals; use particle swarm optimization algorithm to optimize energy distribution paths, with the lowest energy loss as the optimization goal; use LSTM time series prediction model combined with production planning information to predict energy demand and cost under different future production plans and market conditions; specifically include the following steps:

[0148] Step S5-1: Use genetic algorithm to optimize equipment operating parameters:

[0149] Determine the decision variables and encoding method:

[0150] Analyze the key operating parameters that affect the energy consumption and energy utilization efficiency of the equipment, such as the equipment's operating frequency, power setting, operating time, etc., and determine these parameters as decision variables;

[0151] Select appropriate encoding methods according to the characteristics of decision variables, such as binary encoding or real number encoding; binary encoding is suitable for discrete decision variables, and the value range of each decision variable is divided into several intervals, and each interval is represented by a binary number; real number encoding directly represents the value of the decision variable with a real number, and is suitable for continuous decision variables. In this method, a suitable encoding method can be selected according to the actual properties of the equipment operation parameters, and the equipment operation parameter encoding adopts binary encoding or real number encoding (where binary encoding converts the parameter value into a binary digital string, which is convenient for genetic operations in genetic algorithms; real number encoding directly uses the actual value of the parameter for operation, and can more accurately represent the parameter in some cases);

[0152] Construct a fitness function:

[0153] The fitness function is constructed with the lowest energy consumption cost per unit product and the highest energy utilization efficiency as the optimization goal. (in, is the fitness function value, which is used to evaluate the quality of an individual (i.e., a set of equipment operating parameter combinations); and It is a weight coefficient used to balance the importance of unit product energy consumption cost and energy utilization efficiency in the optimization goal. It can be set according to the actual situation of the mine and the management focus. represents the weight of unit product energy consumption cost in the fitness function, Represents the weight of energy utilization efficiency in the fitness function; The energy consumption cost per unit product is an important indicator for measuring the economic efficiency of equipment operation; Energy efficiency, which reflects the degree to which the equipment effectively utilizes energy);

[0154] Initialize the population:

[0155] According to the coding method and value range of decision variables, the initial population is randomly generated; the population size The number of individuals is determined according to the complexity of the problem and the computing resources, usually between dozens and hundreds. Each individual represents a set of equipment operating parameter combinations. The individuals in the initial population should be distributed as evenly as possible in the solution space to increase the possibility of finding the global optimal solution.

[0156] Genetic Operations:

[0157] Selection operation: Roulette selection operation (in, For individuals The probability of being selected, For individuals The fitness function value of each individual is calculated by comparing the fitness function value of each individual to the total fitness function value of all individuals in the population to determine the probability of the individual being selected. The higher the fitness, the greater the probability of the individual being selected, so that excellent individuals have more opportunities to participate in subsequent genetic operations);

[0158] Crossover operation: select single-point crossover operation or multi-point crossover operation; for binary coding, single-point crossover operation is to randomly select a crossover point in the code string of two parent individuals, exchange some genes after the crossover point, and generate two new offspring individuals; multi-point crossover operation is to select multiple crossover points for gene exchange; for real number coding, crossover operation can be to generate offspring individuals by weighted average of the decision variables of two parent individuals; crossover probability Generally, it is between 0.6 and 0.9, which controls the frequency of crossover operations. An appropriate crossover probability helps introduce new gene combinations into the population and increase the diversity of solutions.

[0159] Mutation operation: basic bit mutation operation is used; for binary coding, some bits in the individual code string are randomly selected for inversion operation (i.e. 0 becomes 1, 1 becomes 0); for real number coding, the value of one or some decision variables is randomly disturbed within the value range of the decision variable; mutation probability Generally small, between 0.001-0.1. The mutation operation is mainly used to prevent the algorithm from falling into the local optimal solution. By introducing a small amount of random mutation, the algorithm has the opportunity to jump out of the local optimal area and explore a wider solution space.

[0160] Iterative Optimization:

[0161] Iterative search is performed according to the predetermined number of iterations; in each iteration, selection, crossover and mutation operations are performed in sequence to generate a new population, and then the fitness function value of each individual in the new population is calculated; as the iteration proceeds, the individuals in the population gradually evolve towards the optimal solution, and the fitness function value continues to increase; when the predetermined number of iterations is reached or other termination conditions are met (such as the change in the fitness function value for several consecutive generations is less than the set threshold), the iteration is stopped and the optimal equipment operation parameter combination is output, which is the parameter setting that enables the equipment performance to reach a better state under the current optimization goal;

[0162] Step S5-2: Optimize the energy distribution path using particle swarm optimization algorithm:

[0163] Coding and initialization:

[0164] Encode the energy distribution path into the particle position in the particle swarm optimization algorithm (where the particle position represents the distribution ratio or distribution method of energy between different equipment, process links or energy storage facilities, and the actual energy distribution path is converted into a numerical form that can be processed by the algorithm through encoding);

[0165] Randomly initialize the position and velocity of the particle swarm; the particle swarm size M is determined according to the problem size and computing resources, which is generally similar to the population size in the genetic algorithm; the position vector of each particle represents a possible energy distribution path, and the velocity vector determines the moving direction and speed of the particle in the search space; during initialization, the position and velocity are randomly generated within a reasonable range of values ​​to ensure that the particles have a certain degree of dispersion in the solution space to cover more possible solutions;

[0166] Determine the optimization objective and fitness function:

[0167] The fitness function is constructed with the minimum energy loss as the optimization goal; the fitness function value is directly related to the energy loss amount. The smaller the energy loss, the smaller the fitness function value (the specific function form can be determined according to the calculation method and model of energy loss. For example, the energy loss amount can be directly used as the fitness function value, or some transformations can be used to make the fitness function easier to handle and converge during the optimization process);

[0168] Set the algorithm parameters:

[0169] Setting Inertia Weight , learning factor and ; Inertia weight Controls the influence of the particle's previous velocity on the current velocity. It is usually set between 0 and 1. Good for global search, smaller Good for local search; learning factor and Determine the influence of the particle's own optimal position and the group's optimal position on the particle velocity update, usually Left and right, they jointly guide the particles to move towards the optimal solution in the search space;

[0170] Iteratively update particle states:

[0171] Update the formula based on particle velocity , in, and Particles In the Second and The speed of iterations, is the inertia weight, and is the learning factor, and for A random number, For particles In the The individual optimal position of the iteration, For the group The global optimal position of the iteration, and Particles In the Second and The position of the iteration;

[0172] For each particle, calculate the fitness function value corresponding to its new position and compare it with the individual historical optimal fitness value and the group historical optimal fitness value; if the fitness value of the new position is better, update the individual optimal position and the group optimal position (if the fitness value of the particle's current position is less than its individual historical optimal fitness value, update the current position to the individual optimal position; if the fitness value of the particle's current position is less than the group historical optimal fitness value, update the current position to the group optimal position);

[0173] Termination condition judgment:

[0174] Repeat step 4 until the termination condition is met; the termination condition can be reaching the predetermined maximum number of iterations, or the optimal fitness value of the group has no obvious improvement in several consecutive iterations (i.e., the change is less than the set threshold); when the termination condition is met, the optimal position of the group is output, and the energy allocation path corresponding to the position is the optimized energy allocation scheme, which can minimize or nearly minimize the energy loss;

[0175] Step S5-3: Predict energy demand and cost using LSTM time series prediction model:

[0176] Data preparation and preprocessing:

[0177] Collect historical energy cost data, production data (such as output, equipment operating time, processing volume, etc.) and influencing factor data (such as market energy price fluctuations, seasonal factors, production process adjustments, etc.) as input variables (historical energy cost data is used to reflect the relationship between past energy consumption and costs, and is the basis for predicting future costs; production data is closely related to energy consumption. By analyzing the historical relationship between production data and energy costs, the impact of production activities on energy demand and costs can be explored; influencing factor data covers the potential impact of external environment and internal production changes on energy costs, which helps to improve the accuracy of predictions);

[0178] Before inputting into the model, production planning information (including planned mining volume , Planned beneficiation volume , Planned equipment operating time Normalization is to map data of different magnitudes to a specific interval, such as [0,1] or [-1,1], in order to eliminate the impact of the dimensions between the data, so that the model can converge more stably and quickly during training, thereby improving the prediction performance of the model.

[0179] Construct the LSTM model structure:

[0180] Determine the input layer, hidden layer and output layer structure of the LSTM model; the number of input layer nodes is determined according to the number of input variables, and each input variable corresponds to an input node; the hidden layer can be set to one or more layers, and the number of hidden layer nodes is determined by experiments or experience, generally between dozens and hundreds, and the hidden layer is used to extract features and patterns in the data; the number of output layer nodes is determined according to the prediction target. In this method, if energy demand and cost are predicted at the same time, the output layer can be set to have two nodes, which output energy demand prediction values ​​and cost prediction values ​​respectively (the LSTM unit plays a key role in memorizing and processing time series information in the model, and controls the forgetting, updating and output of information through the gating mechanism, so that data with time series characteristics can be effectively processed);

[0181] Model parameter initialization:

[0182] Initialize the weight matrix and bias vector of the LSTM unit; including the input gate weight matrix , forget gate weight matrix , output gate weight matrix , candidate memory unit weight matrix and the corresponding bias vector , , , (These weight matrices and bias vectors are important parameters in the LSTM model. They determine how the model processes input data and its learning ability. The selection of initialization values ​​will affect the training effect and convergence speed of the model. Random initialization is usually used. During the training process, these parameters are continuously adjusted through the back propagation algorithm to enable the model to better fit the data).

[0183] Model training:

[0184] The historical data is divided into a training set and a validation set. The training set is used to learn the model parameters, and the validation set is used to evaluate the performance of the model and prevent overfitting.

[0185] For each sample in the training set, input the data in sequence according to the time step, and calculate the LSTM unit at time Input gate (in, is the value of the input gate, which determines how much input information can enter the memory unit at the current moment. is the sigmoid function, which compresses the calculation result to between 0 and 1. is the input gate weight matrix, is the hidden layer output at the previous moment, Input for the current moment, is the input gate bias vector), forget gate ( is the forget gate weight matrix, is the forget gate bias vector), output gate (The output gate determines how much information in the memory unit can be output to the hidden layer at the current moment, is the output gate weight matrix, is the output gate bias vector), candidate memory cells (Candidate memory cells are used to generate new information that needs to be updated to the memory cell at the current moment. is the candidate memory unit weight matrix, is the candidate memory unit bias vector), and then updates the memory unit (in, Represents element-wise multiplication, memory unit It combines the forgetting of the memory unit information at the previous moment and the updating of the new information at the current moment), and finally calculates the hidden layer output ;

[0186] According to the error between the predicted output and the actual output (such as mean square error MSE or mean absolute error MAE, etc.), the back propagation algorithm is used to calculate the gradient, and the weight matrix and bias vector of the LSTM unit are updated through the gradient descent method or other optimization algorithms (such as the Adam optimization algorithm); during the training process, the model parameters are continuously adjusted to gradually reduce the error until the predetermined training stop condition is reached (such as the number of training rounds reaches the set value, the error of the verification set no longer decreases or reaches the minimum error threshold, etc.);

[0187] Model prediction and evaluation:

[0188] Use the trained LSTM model, combined with production planning information (normalized), to predict energy demand and cost under different future production plans and market conditions; compare and evaluate the prediction results with actual data (if some future actual data is available for comparison) or the results of other benchmark prediction methods. Evaluation indicators may include root mean square error (RMSE), mean absolute percentage error (MAPE), etc., to measure the accuracy and reliability of model predictions; based on the evaluation results, if the model performance is not ideal, further adjustments can be made to the model structure, parameters, or the amount of data to improve the model and improve the prediction accuracy, thus providing a more valuable reference for energy management and production decisions in the mine;

[0189] Step S6: Visualization and decision support: Develop a visualization interface to display equipment energy consumption, process energy consumption, and energy cost data; automatically generate an energy management decision support report based on cost accounting and optimization analysis results and push it to mine managers; the specific steps are as follows:

[0190] Step S6-1: Visualization interface development and data display:

[0191] Select visualization tools and techniques:

[0192] According to the actual needs and technical environment of the mine, choose appropriate visualization tools or development frameworks, such as Matplotlib, Seaborn, Plotly and other libraries in Python, or use professional commercial visualization software (such as Tableau, PowerBI, etc.); these tools provide a variety of chart types and interactive functions to meet the visualization needs of different data;

[0193] Data connection and acquisition:

[0194] Establish a connection between the visualization interface and the data storage system (such as a relational database, distributed file system, etc.), and obtain relevant data such as equipment energy consumption, process energy consumption, energy cost, etc. from the data storage system (ensure the real-time or timeliness of the data, and set a reasonable data acquisition mechanism according to the frequency of data update, such as regularly querying the latest data from the database);

[0195] Graphic design and layout:

[0196] Design appropriate chart types based on data characteristics and analysis purposes; for equipment energy consumption data, a bar chart can be used to show the energy consumption comparison of different equipment over a period of time, or a line chart can be used to show the energy consumption trend of a single device over time; process energy consumption can be presented in the form of a flow chart combined with energy consumption data annotation to intuitively show the energy consumption distribution of each process link; energy cost data is suitable for using a pie chart to show the proportion of different energy costs in the total cost, or a bar chart to compare energy cost changes in different time periods;

[0197] Design the layout of the visualization interface and arrange different charts reasonably on the interface to ensure that the information is clear and easy to understand. You can add elements such as titles, axis labels, and legends to enhance the readability of the charts. At the same time, consider setting up interactive functions, such as hovering the mouse to display data details, zooming and panning charts to view different time periods or data ranges, etc., to facilitate mine managers to explore and analyze data;

[0198] Data visualization implementation:

[0199] Use the selected visualization tool to visualize the acquired data according to the designed chart type and layout; aggregate, group and transform the data according to the time dimension of the data (such as day, week, month, year, etc.) and classification dimensions such as equipment and process links to meet the display requirements of different charts; for example, summarize the daily equipment energy consumption data into monthly data to draw a monthly energy consumption trend chart;

[0200] Optimize the visualization effect and adjust visual elements such as color, font, line style, etc. to make the chart beautiful and in line with the visual specifications of the mining company; ensure that the chart can be displayed normally and has good readability on different devices (such as computer monitors, tablets, mobile phones, etc.) to facilitate managers to view data anytime and anywhere;

[0201] Step S6-2: Decision support report generation and delivery:

[0202] Report template design:

[0203] Design an energy management decision support report template based on the content and format requirements of the cost accounting and optimization analysis results; the report template should include a cover, table of contents, summary, detailed analysis content, conclusions and suggestions, etc.; in the detailed analysis content, it should include cost accounting results (such as details of each energy cost, total cost change trend, etc.), optimization analysis results (such as comparison of equipment operating parameters before and after optimization, energy distribution path optimization effect, future energy demand and cost forecast, etc.), and combine charts and data for intuitive display;

[0204] Report content generation:

[0205] Extract relevant data of cost accounting and optimization analysis from the data storage system, and fill the data into the corresponding chapters and charts according to the format requirements of the report template; automatically generate text descriptions, such as summarizing the reasons for cost changes, the implementation effects of optimization measures, and the analysis of future energy trends. The text descriptions should be concise, logical, and able to accurately convey key information;

[0206] Add data interpretation and analysis to the report to help mine managers understand the meaning and potential problems behind the data; for example, compare the energy consumption of different equipment with the industry average, analyze the impact of changes in energy cost ratio on corporate profits, etc., to provide managers with in-depth information needed for decision-making;

[0207] Report push settings:

[0208] Determine the method and frequency of report push, such as email, SMS or internal system message push (select the appropriate push method according to the working habits and information access convenience of mine managers to ensure that the report can be delivered to managers in a timely manner); set the push time, such as pushing the energy management report of the previous week every Monday morning, or pushing the detailed analysis report of the previous month at the beginning of each month, so that managers can regularly obtain the latest energy management information;

[0209] Configure the push system to ensure that the report can be accurately sent to the designated management account or device; for email push, set the email subject, text format and attachments (such as PDF or Excel format files of the report); for SMS push, control the length of the SMS and highlight key information; for internal system message push, ensure that the message is displayed in a prominent position in the mine management system and provide a convenient link to view the full text of the report;

[0210] Report Feedback and Improvements:

[0211] Collect feedback from mine managers on the report to understand their satisfaction with the report content, format and delivery method, as well as their needs for further improvement; based on the feedback, optimize and adjust the report template, generated content and delivery settings, and continuously improve the quality and practicality of decision support reports to better serve the mine's energy management decision-making process; for example, if managers want to add specific analysis dimensions or chart types to the report, modify the report generation system in a timely manner to meet user needs;

[0212] Step S7: Feedback and continuous optimization: Regularly collect actual energy consumption, cost and equipment operation data and compare them with system prediction results, collect information on equipment failures, production process adjustments, and market energy price fluctuations, and optimize and adjust the energy loss model, cost accounting model, and optimization algorithm based on the comparison and collection information. Specifically, the following steps are included:

[0213] Step S7-1: Data collection and comparison:

[0214] Regular data collection:

[0215] Set a fixed data collection cycle, such as weekly, monthly or quarterly, and determine the appropriate cycle based on the actual production situation of the mine and the frequency of data changes (the collection cycle should be able to reflect the dynamic changes of energy consumption and equipment operation in a timely manner, and not increase unnecessary workload and system burden due to too frequent data collection); in each collection cycle, obtain actual energy consumption, cost and equipment operation data from various monitoring equipment in the mine (such as smart meters, fuel flow meters, sensors, etc.) and cost accounting systems to ensure the accuracy and completeness of the data (conduct a preliminary quality check on the collected data, such as checking the rationality and completeness of the data, eliminating obviously erroneous or missing data points, and using appropriate interpolation methods to supplement a small amount of missing data);

[0216] Compared with the predicted results:

[0217] Extract the forecast results of energy consumption, cost and equipment operation of the corresponding period from the system (these forecast results are generated by the forecast model in the previous step, such as the forecast of energy demand and cost by the LSTM time series forecast model, and the energy consumption forecast after the equipment operation parameters are optimized based on the optimization algorithm); compare the actual collected data with the forecast results one by one, and calculate the deviation values ​​of each data, such as absolute deviation, relative deviation, etc. (absolute deviation = actual value - forecast value, relative deviation = (actual value - forecast value) / forecast value × 100%, through the calculation of the deviation value, you can intuitively understand the degree of deviation between the forecast result and the actual situation); analyze the causes of the deviation, which may include the limitations of the forecast model, the changes in the external environment are not fully considered (such as sudden market energy price fluctuations, abnormal climatic conditions affect equipment operation efficiency, etc.), equipment aging or failure leads to abnormal increase in energy consumption, etc. (the analysis of the causes of the deviation requires comprehensive consideration of multiple factors, combined with the actual production situation of the mine and the equipment operation status for in-depth exploration);

[0218] Step S7-2: Information collection:

[0219] Equipment failure information collection:

[0220] Establish an equipment failure monitoring and reporting mechanism, and obtain equipment failure information in a timely manner through the equipment's own fault diagnosis system (if any), feedback from on-site operators, and regular equipment inspection records (equipment failure information should include details such as the time of failure, equipment name, failure type, and failure severity, so as to accurately analyze the impact of the failure on energy consumption and production); classify and sort out equipment failures, count the frequency and impact range of different types of failures, and analyze the relationship between failures and increased energy consumption (for example, the failure of certain key equipment may cause the entire production process to stagnate or operate inefficiently, thereby significantly increasing energy consumption. Analyzing this relationship can provide a targeted basis for subsequent optimization and adjustment);

[0221] Production process adjustment information collection:

[0222] Maintain close communication with the mine production department and obtain relevant information on production process adjustments in a timely manner, such as process improvements, introduction of new equipment or new technologies, changes in production tasks, etc. (understand the background, purpose and specific implementation content of production process adjustments, as well as changes in production processes and equipment operation modes before and after adjustments); record the time nodes of production process adjustments and analyze their impact on energy consumption, costs and equipment operation (production process adjustments may change the load characteristics of equipment, energy demand structure and production efficiency, thereby affecting energy loss and cost accounting. These impacts need to be accurately evaluated so that process adjustment factors can be considered in model optimization);

[0223] Market energy price fluctuation information collection:

[0224] Pay attention to the dynamics of the energy market, collect information on market energy price fluctuations by subscribing to energy price information platforms, keeping in touch with energy suppliers or referring to industry reports (obtain price change data of the main energy sources used by mines, such as electricity, fuel oil, and coal, including the extent of price increases or decreases, time periods, and price trend forecasts); analyze the direct impact of market energy price fluctuations on mine energy costs, as well as the indirect impact (such as price fluctuations may prompt mines to adjust energy procurement strategies and optimize energy use structures, thereby affecting equipment operating parameters and production process arrangements), and provide a basis for market price factors for the optimization of cost accounting models;

[0225] Step S7-3: Model and algorithm optimization and adjustment:

[0226] Energy loss model optimization:

[0227] According to the data comparison results and the collected information, the energy loss model is optimized and adjusted; if it is found that the actual energy loss deviates greatly from the model prediction and is mainly caused by changes in the equipment operating status or external environmental factors, it may be necessary to re-evaluate and correct the parameters in the equipment energy consumption model (such as the relevant parameters in the power characteristic function, load impact function, and environmental impact correction function) to more accurately reflect the energy consumption characteristics of the equipment under actual working conditions (for example, equipment aging causes power to decrease, and the parameters in the power characteristic function need to be adjusted accordingly; the ambient temperature is higher than the model assumption value for a long time, which affects the heat dissipation of the equipment and thus increases energy consumption, and the environmental impact correction function needs to be optimized); for the process energy consumption model, if the energy consumption relationship of each link of the equipment changes after the production process is adjusted, the equipment energy consumption accumulation relationship and auxiliary energy consumption calculation method in the process energy consumption model are updated in time (to ensure that the process energy consumption model can adapt to the energy consumption after the process changes); at the same time, according to the changes in the loss proportion of each part in the energy loss decomposition model, the loss calculation method and weight distribution of the equipment, process link, and energy type in the model are optimized to make the energy loss model closer to the energy loss in the actual production process;

[0228] Cost accounting model optimization:

[0229] For the cost accounting model, optimize it based on the comparison results of market energy price fluctuation information and actual cost data; if energy prices fluctuate frequently and have a significant impact on costs, add price fluctuation adjustment factors to the cost accounting model so that cost calculation can more timely and accurately reflect market price changes (for example, establish a dynamic correlation model between electricity costs and real-time electricity price fluctuations, and update electricity cost calculations in a timely manner according to electricity price adjustments); combine the analysis of the impact of production process adjustments on costs, adjust the cost calculation parameters related to production processes in the cost accounting model (such as the energy consumption cost allocation coefficient of different process links, the correlation between equipment maintenance costs and process changes, etc.), to ensure that cost accounting can accurately reflect actual cost changes after process adjustments; at the same time, optimize the calculation methods of other cost items in the cost accounting model (such as considering the impact of equipment failure repair costs on total costs, incorporating them into the cost accounting model to make cost accounting more comprehensive);

[0230] Optimization algorithm improvements:

[0231] According to the data comparison and analysis results, evaluate the performance of genetic algorithms, particle swarm optimization algorithms, etc. in the process of optimizing equipment operating parameters and energy distribution paths; if the optimization results fail to achieve the expected results or new optimization needs are found in actual production, improve the optimization algorithm; for example, adjust the encoding method, crossover probability, mutation probability and other parameters in the genetic algorithm to improve the algorithm's search efficiency and global optimization ability (by experimentally comparing the performance of the algorithm under different parameter settings, find the most suitable parameter combination for the actual situation of the mine); for the particle swarm optimization algorithm, optimize the inertia weight, learning factor value strategy, or improve the particle position and speed update formula, so that the algorithm can converge to the optimal solution faster and more accurately (according to the performance of the algorithm in actual application, the results are summarized). At the same time, new optimization algorithms or technologies should be continuously introduced (such as research on the application potential of deep learning algorithms in energy management), compared and integrated with existing algorithms, to explore optimization methods that are more suitable for mine energy loss balance management, and to improve mine energy utilization efficiency and cost control level; in the process of optimization and adjustment, new models and algorithms should be fully tested and verified to ensure their stability and effectiveness, and to avoid new problems or errors caused by optimization and adjustment; historical data backtesting, simulation and small-scale pilot applications can be used for verification, and the models and algorithms can be further improved according to the verification results to form a virtuous cycle of continuous optimization and continuously improve the scientificity and accuracy of mine energy management.

[0232] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A method for cost statistics based on mine energy loss balance, characterized by: The following steps are involved: Step S1, data collection and transmission: Install smart meters as power monitors at the power equipment in the mine, set fuel flow meters and equip fuel level sensors on fuel consumption equipment, deploy temperature sensors, humidity sensors, pressure sensors in the mine area, and install operating status sensors on the equipment, and transmit the data collected by all sensors to the data processing center through wired and wireless networks; Step S2: Data preprocessing and storage: In the data processing center, anomaly detection methods are used to remove abnormal data points based on data statistical laws and equipment operation logic, and interpolation algorithms are used to supplement missing data. Data is stored in relational databases and distributed file systems according to equipment type, production process, energy type, and time dimension. Step S3: Energy loss model construction: for each equipment in the mine, the equipment energy consumption model is constructed by comprehensively considering the equipment power characteristics, load influence, and environmental factors; the mine production process is analyzed, and the process energy consumption model is constructed based on the accumulated energy consumption of equipment in each link and the auxiliary energy consumption calculation; the energy loss cost is obtained by subtracting the effective output energy consumption from the total energy input of the mine, and an energy loss decomposition model is established according to the equipment, process link, and energy type; Step S4: Cost accounting: Calculate the electricity cost according to the electricity pricing method of the power company and the data of the power monitor, calculate the fuel cost according to the fuel flow meter data and the unit price of fuel, summarize the total energy cost of the mine and analyze the proportion of each energy cost; Step S5: Optimization analysis: Use genetic algorithm to optimize equipment operating parameters, with the lowest unit product energy consumption cost and the highest energy utilization efficiency as the optimization goal; use particle swarm optimization algorithm to optimize the energy distribution path, with the minimum energy loss as the optimization goal; Use the LSTM time series forecasting model combined with production planning information to predict energy demand and cost under different future production plans and market conditions; Step S6: Visualization and decision support: Develop a visualization interface to display equipment energy consumption, process energy consumption, and energy cost data; Automatically generate energy management decision support reports based on cost accounting and optimization analysis results and push them to mine managers; Step S7: Feedback and continuous optimization: Regularly collect actual energy consumption, cost and equipment operation data and compare them with system prediction results, collect information on equipment failures, production process adjustments, and market energy price fluctuations, and optimize and adjust energy loss models, cost accounting models, and optimization algorithms based on the comparison and collection of information.

2. The method for cost statistics based on mine energy loss balance according to claim 1, characterized in that: The anomaly detection method is based on The anomaly detection method based on the principle is , calculate its mean and standard deviation , like or , it is determined as abnormal data and removed; the interpolation algorithm is a cubic spline interpolation algorithm.

3. The method for cost statistics based on mine energy loss balance according to claim 1, characterized in that: The device energy consumption model is constructed as follows: Assume that the device power characteristic function is , the load influence function is , the environmental impact correction function is , then the equipment energy consumption model in, is the time variable, is the load variable, is the environment variable vector, Indicates that the device is , and The energy consumption under the conditions, the environmental variable vector includes temperature, humidity, pressure, and altitude; the energy consumption model of the process flow is constructed as follows: The energy consumption of the device is , the auxiliary energy consumption is , then the process energy consumption model , represents the total energy consumption of the process, Indicates the number of equipment in the process flow.

4. The method for cost statistics based on mine energy loss balance according to claim 1 is characterized by: The calculation method of the electricity cost is: Assume that the unit price of electricity is The total power consumption is , electricity cost , when the electricity pricing method includes basic electricity charges , peak and valley electricity charges, power factor adjustment electricity charges The peak-valley electricity fee includes the peak-period unit price. , Peak power consumption , Off-peak hours unit price , off-peak hours , time period unit price , power consumption during normal periods , electricity cost ; The fuel cost accounting method is: Assume the fuel unit price is The total fuel consumption is , fuel costs The total energy cost is calculated as ,in, For other energy costs.

5. The method for cost statistics based on mine energy loss balance according to claim 1, characterized in that: The genetic algorithm is applied in the following way: the equipment operation parameter coding adopts binary coding or real number coding, and the fitness function is constructed with the lowest unit product energy consumption cost and the highest energy utilization efficiency as the optimization goal. ,in, and is the weight coefficient, is the energy consumption cost per unit product, For energy efficiency, select the operation by roulette wheel , single-point crossover operation or multi-point crossover operation, basic bit mutation operation, iteratively search for the optimal device operating parameter combination according to a predetermined number of iterations, wherein, For individuals The probability of being selected, For population size.

6. The method for cost statistics based on mine energy loss balance according to claim 5 is characterized by: The particle swarm optimization algorithm is applied as follows: the energy distribution path is encoded as the particle position in the particle swarm optimization algorithm, the minimum energy loss is taken as the optimization goal, and the particle speed update formula is: , the particle position update formula is ,in, and Particles In the Second and The speed of iterations, is the inertia weight, and is the learning factor, and for A random number, For particles In the The individual optimal position of the iteration, For the group The global optimal position of the iteration, and Particles In the Second and The position of the iteration, initialize the position and velocity of the particle swarm, and set the inertia weight Learning Factor and .

7. The method for cost statistics based on mine energy loss balance according to claim 6 is characterized by: The LSTM time series prediction model is applied in the following way: historical energy cost data, production data and influencing factor data are used as input variables. Before inputting the model, production planning information including planned mining volume , Planned beneficiation volume , Planned equipment operating time Perform normalization processing and assume that the LSTM unit is at time Input gate ,in, is the input gate weight matrix, is the hidden layer output at the previous moment, Input for the current moment, is the input gate bias vector, is the sigmoid function, the forget gate ,in, is the forget gate weight matrix, is the bias vector of the forget gate, the output gate ,in, is the output gate weight matrix, is the output gate bias vector, candidate memory unit ,in, is the candidate memory unit weight matrix, is the candidate memory unit bias vector, memory unit ,in, Represents element-wise multiplication, the hidden layer output , by adjusting the LSTM unit weights and biases through learning and training of historical data, and combining production planning information to predict energy demand and cost under different future production plans and market conditions.

8. The method for cost statistics based on mine energy loss balance according to claim 1, characterized in that: The visualization interface display forms include bar charts, line charts, and pie charts; the decision support report push methods include email, mobile phone text messages, or system internal message push.

Images