Lithium ore hopper energy-saving conveying system based on intelligent self-adaptive control

Abstract

The invention relates to the technical field of lithium ore conveying, and discloses a lithium ore hopper energy-saving conveying system based on intelligent self-adaptive control. The system comprises a data acquisition module, a self-adaptive control module, an energy-saving optimization module and an execution adjustment module. The data acquisition module acquires multi-dimensional operation data of the lithium ore hopper through a distributed sensor network, and a hopper conveying state database is constructed after standardization processing; the adaptive control module performs dynamic feature extraction on the database, generates a real-time control parameter matrix, establishes an association rule graph of each parameter, and constructs an intelligent adjustment decision model based on the association rule graph; an energy-saving optimization module extracts a key energy consumption feature vector from the real-time control parameter matrix, calculates an energy consumption influence coefficient of each operation parameter, and screens out an optimal energy-saving control sequence in combination with an intelligent adjustment decision model; and the execution adjustment module drives an execution mechanism according to the optimal energy-saving control sequence, and self-adaptive adjustment of hopper conveying parameters is completed.

Description

Technical Field

[0001] The present invention relates to the technical field of lithium ore transportation, and in particular to an energy-saving lithium ore hopper transportation system based on intelligent adaptive control. Background Art

[0002] In the lithium mining and processing process, the lithium ore hopper conveying system is a critical equipment cluster connecting the mining and beneficiation stages. Its operating status is directly related to the continuity and economic efficiency of lithium ore processing. Currently, traditional lithium ore hopper conveying systems have exposed many problems in practical applications that need to be solved.

[0003] Traditional systems rely heavily on preset parameters or manual adjustments based on experience, making them incapable of responding to the dynamic changes in lithium ore properties. Physical properties of lithium ore, such as particle size distribution, moisture content, and bulk density, fluctuate significantly with changes in mining area and depth. For example, ore mined during the rainy season experiences increased moisture content, making it prone to clumping, while ore mined during the dry season exhibits uneven particle size and significant differences in fluidity. Under fixed parameter control, the operating rhythm of conveying equipment doesn't match the actual ore characteristics, often leading to an imbalance between conveying speed and ore supply. When the ore supply suddenly increases, if the conveying speed isn't increased in time, the hopper can easily become clogged, requiring shutdown and cleaning, disrupting the production process. When the ore supply decreases, maintaining a high conveying speed can cause the equipment to idle, wasting energy unnecessarily.

[0004] At the data collection level, existing systems often use single-point sensors or limited-area monitoring. These systems collect data in limited dimensions, such as rotational speed and voltage, making it difficult to fully capture key information such as pressure distribution within the hopper, ore flow patterns, and equipment vibration frequency. This one-sided data collection model prevents operators from understanding the hopper's true operating status in real time, and they lack the ability to predict potential abnormal operating conditions (such as increased local wear or abnormally high bearing temperatures). Consequently, remedial measures are often implemented only after a fault has occurred, increasing equipment maintenance costs and downtime.

[0005] The lack of adaptive adjustment capabilities is another prominent issue. Parameter adjustments in traditional systems often respond with a lag. That is, parameter adjustments are only made through manual intervention or simple feedback mechanisms after a significant equipment anomaly (such as excessive current or conveyor belt deviation) occurs. This passive adjustment model not only fails to prevent the continued development of abnormal operating conditions, but can also cause new problems due to excessive adjustments. For example, a sudden increase in conveyor speed to resolve a blockage can overload the conveyor belt and accelerate the aging of motors and transmission components.

[0006] In terms of energy-saving control, existing technologies often focus on the static optimization of a single parameter, such as simply reducing motor speed or conveyor belt tension, without considering the interrelationships between these parameters. In reality, the energy consumption of a hopper conveying system is the result of the combined effects of multiple parameters, including motor power, conveyor belt speed, hopper inclination, and vibration frequency. Adjusting a single parameter can lead to a compensatory increase in the energy consumption of other parameters. For example, while reducing motor speed can reduce direct energy consumption, it may cause the ore to remain in the hopper for a longer time, increasing friction loss on the hopper's inner wall and ultimately increasing overall energy consumption.

[0007] Existing systems for energy consumption monitoring often rely on isolated meters or sensors, making it impossible to establish a quantitative correlation between energy consumption and operating parameters, making it difficult to identify key energy consumption points. This results in a lack of targeted energy-saving measures, often relying on a "one-size-fits-all" approach to extensive management. This approach fails to dynamically optimize energy consumption based on real-time operating conditions, resulting in significant energy waste. Amidst increasing energy price volatility and increasingly stringent environmental regulations, this high-energy-consumption, low-adaptability transportation model not only drives up production costs for lithium ore processing but also runs counter to the development of green mines. Summary of the Invention

[0008] The purpose of the present invention is to provide an energy-saving lithium ore hopper conveying system based on intelligent adaptive control to solve the problems raised in the above background technology.

[0009] To achieve the above objectives, the present invention provides an energy-saving lithium ore hopper conveying system based on intelligent adaptive control, the system comprising: A data acquisition module is used to collect multi-dimensional operating data of the lithium ore hopper through a distributed sensor network, standardize the multi-dimensional operating data, and build a hopper conveying status database; An adaptive control module is used to extract dynamic features from the hopper conveying state database, generate a real-time control parameter matrix, establish an association rule graph for each parameter in the real-time control parameter matrix, and construct an intelligent adjustment decision model based on the association rule graph; An energy-saving optimization module, configured to extract key energy consumption feature vectors from the real-time control parameter matrix, calculate the energy consumption impact coefficient of each operating parameter based on the key energy consumption feature vectors, and select the optimal energy-saving control sequence in combination with the intelligent adjustment decision model; The execution adjustment module is used to drive the actuator to complete the adaptive adjustment of the hopper conveying parameters according to the optimal energy-saving control sequence.

[0010] Preferably, the multi-dimensional operation data of the lithium ore hopper includes material property data, equipment operation data, energy consumption monitoring data, environmental perception data and actuator status data; the material property data includes particle size distribution data, moisture content data, bulk density data and hardness data of the lithium ore.

[0011] Preferably, standardizing the multi-dimensional operation data to construct a hopper conveying status database specifically includes: Eliminating outliers and filling missing values ​​in the multidimensional operating data to obtain preprocessed operating data; Perform dimension unification and time-series alignment on the pre-processed operating data to generate standardized feature data; All standardized feature data are structured and stored according to the time dimension and space dimension to build a hopper conveying status database.

[0012] Preferably, extracting dynamic features from the hopper conveying state database to generate a real-time control parameter matrix specifically includes: A sliding window algorithm is used to extract segmented features from the time series data in the hopper conveying state database to obtain local feature vectors for each time period; A real-time control parameter matrix is ​​constructed based on the local feature vectors of each time period and the preset control parameter thresholds.

[0013] Preferably, establishing an association rule diagram of each parameter in the real-time control parameter matrix specifically includes: Selecting a target control parameter from the real-time control parameter matrix and acquiring all associated parameters related to the target control parameter; Calculate the Pearson correlation coefficient between the target control parameter and each associated parameter to determine the parameter association strength; An association rule subgraph of the target control parameters is constructed according to the parameter association strength and the preset association threshold, and an association rule graph of the real-time control parameter matrix is ​​obtained by integrating the association rule subgraphs of all target control parameters; When selecting target control parameters in the real-time control parameter matrix, operating parameters directly related to energy consumption indicators are preferably selected as target control parameters. The operating parameters directly related to energy consumption indicators include motor operating power, conveyor belt operating speed and gate opening degree.

[0014] Preferably, constructing an intelligent adjustment decision model based on the association rule graph specifically includes: Extracting key influencing factors of each control parameter from the association rule graph and determining weight coefficients of each key influencing factor; Establish a multi-objective optimization model based on the weight coefficients of key influencing factors and the preset adjustment objective function; Solving the multi-objective optimization model by using a particle swarm optimization algorithm to generate an intelligent adjustment decision model; When extracting the key influencing factors of each control parameter from the association rule graph, the importance score of each feature is calculated by the random forest algorithm, and features with scores higher than a preset threshold are selected as key influencing factors. The preset threshold is determined based on the average feature contribution of historical adjustment data.

[0015] Preferably, calculating the energy consumption impact coefficient of each operating parameter according to the key energy consumption characteristic vector specifically includes: Obtain the historical energy consumption benchmark value and current energy consumption measured value of each operating parameter; Calculate the deviation rate between the current energy consumption measured value and the historical energy consumption benchmark value, combined with the characteristic contribution of the key energy consumption feature vector; The energy consumption impact coefficient of each operating parameter is determined by the weighted product of the deviation rate and the characteristic contribution.

[0016] Preferably, screening out the optimal energy-saving control sequence in combination with the intelligent adjustment decision model specifically includes: Input the energy consumption impact coefficient of each operating parameter into the intelligent regulation decision model to generate multiple sets of candidate control sequences; Perform energy consumption simulation and transmission efficiency evaluation on each set of candidate control sequences to obtain a comprehensive performance score; The candidate control sequence with the highest comprehensive performance score is selected as the optimal energy-saving control sequence; The energy consumption simulation calculation and transportation efficiency evaluation of each set of candidate control sequences specifically include: simulating the energy consumption per unit time under the candidate control sequence through the system dynamics model, and using the discrete event simulation method to evaluate the average speed and throughput rate of material transportation.

[0017] Preferably, driving the actuator to complete the adaptive adjustment of the hopper conveying parameters according to the optimal energy-saving control sequence is to convert the optimal energy-saving control sequence into a control instruction of the actuator, and send it to the corresponding execution unit through the industrial bus to achieve coordinated adjustment of the conveying speed, gate opening and motor frequency.

[0018] Preferably, the actuator includes a variable frequency speed regulating motor, an electric regulating gate and a pressure sensor feedback unit.

[0019] Compared with the prior art, the present invention has the following beneficial effects: This system operates through the collaborative operation of multiple modules, forming a closed-loop control from data collection to execution and adjustment, comprehensively improving various performances in the lithium ore hopper transportation process.

[0020] Leveraging a distributed sensor network, the data acquisition module transcends the limitations of traditional single-point monitoring, enabling simultaneous collection of multi-dimensional operational data, including lithium ore particle size distribution, hopper pressure, conveyor belt tension, motor temperature, and ambient humidity. The standardized hopper conveying status database encompasses all key factors influencing the conveying process, enabling the system to develop a comprehensive and dynamic understanding of operating status, moving beyond local indicators. This comprehensive data support provides a more realistic basis for subsequent control and optimization, avoiding decision-making biases caused by incomplete information.

[0021] The adaptive control module extracts dynamic features from the state database to generate a real-time control parameter matrix that accurately reflects the specific values ​​and changing trends of each parameter under the current operating conditions. The association rule diagram established on this basis clearly shows the intrinsic connection between parameters such as motor speed and conveyor belt tension, hopper inclination and ore flow rate, making the previously isolated parameters an organic whole that is interconnected. The intelligent adjustment decision model constructed based on this association relationship breaks away from the limitations of traditional control that relies on fixed formulas or experience. It can autonomously derive the adjustment direction based on real-time parameter changes, making the control logic more in line with the dynamic changes in the operating conditions. When the ore particle size suddenly becomes coarser, the model can quickly identify this feature and adjust the conveyor belt speed and hopper vibration frequency based on the association rules to avoid conveying obstructions caused by parameter mismatch.

[0022] The energy-saving optimization module focuses on the deep correlation between energy consumption and operating parameters. By extracting key energy consumption feature vectors, it converts abstract energy consumption indicators into quantifiable and relatable feature parameters. The energy consumption impact coefficient of each operating parameter clearly reveals the contribution of different parameters to the overall energy consumption. For example, the energy consumption impact coefficient of the motor speed under high load is significantly higher than that under low load, and the energy consumption impact coefficient of the hopper inclination angle increases significantly when the ore moisture is high. The optimal energy-saving control sequence selected by the intelligent adjustment decision model is not an extreme value adjustment of a single parameter, but a collaborative optimization of multiple parameters to ensure that energy consumption is controlled within a reasonable range while improving transportation efficiency. When the system detects an increase in ore moisture, the optimal sequence will simultaneously adjust the conveyor belt speed to reduce and the hopper vibration frequency to increase, which not only avoids ore sticking and clogging, but also prevents energy consumption from rising due to excessive adjustment of a single parameter.

[0023] The execution and adjustment module drives the actuator according to the optimal energy-saving control sequence, achieving a seamless transition from decision-making to action. Compared to the lag and error inherent in manual adjustments in traditional systems, this automated, adaptive adjustment can complete parameter corrections within milliseconds, ensuring that the hopper's conveying status always matches the ore characteristics and production requirements. When the pressure within the hopper exceeds the normal range, the actuator quickly adjusts the hopper's inclination and the size of the discharge port, maintaining a stable conveying rhythm while avoiding blockages and reducing energy and time waste caused by downtime or failures. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] Figure 1 This is a working principle diagram of the lithium ore hopper energy-saving conveying system based on intelligent adaptive control according to the present invention; Figure 2 A flow chart constructed for the hopper conveying status database; Figure 3 Flowchart for generating real-time control parameter matrix for dynamic feature extraction; Figure 4 Flowchart for screening the optimal energy-saving control sequence. DETAILED DESCRIPTION

[0025] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0026] See also Figure 1 The present invention provides an energy-saving lithium ore hopper conveying system based on intelligent adaptive control, the system comprising: The data acquisition module collects multidimensional operational data from the lithium ore hopper through a distributed sensor network, including a particle size sensor installed at the hopper inlet, a humidity sensor on the hopper's inner wall, a pressure sensor on the conveyor belt, a current sensor in the motor control cabinet, and temperature and humidity sensors around the hopper. Once collected, the multidimensional operational data is standardized to remove redundant information and interference. The processed data is then categorized and stored according to a pre-set data structure to construct a hopper conveying status database, providing data support for subsequent control decisions.

[0027] The adaptive control module draws on data from the hopper conveying status database and uses a dynamic feature extraction algorithm to perform in-depth analysis of the data, identifying key features reflecting the hopper's operating status and generating a real-time control parameter matrix. Based on this, the module analyzes the inherent connections between the parameters in the matrix and establishes an association rule graph to clarify the impact of different parameters on the conveying process. Based on this association rule graph and combined with a machine learning algorithm, an intelligent adjustment decision model is constructed, which can dynamically adjust the control strategy based on real-time data.

[0028] The energy-saving optimization module extracts key eigenvectors related to energy consumption from the real-time control parameter matrix and determines the energy consumption impact coefficient by calculating the impact of each operating parameter on energy consumption. This energy consumption impact coefficient is input into the intelligent regulation decision model, which, through multiple rounds of iterative calculations, selects the optimal energy-saving control sequence that balances transmission efficiency and energy consumption, achieving energy conservation goals.

[0029] The execution and adjustment module receives the optimal energy-saving control sequence, converts it into control instructions that can be recognized by the actuator, and sends it to the corresponding execution unit through the industrial communication protocol, driving the actuator to complete real-time adjustment of parameters such as conveying speed and gate opening, ensuring that the hopper conveying system is always in the optimal operating state.

[0030] Example 1: See Figure 2 Multi-dimensional operational data covers material characteristics, equipment operation data, energy consumption monitoring data, environmental perception data, and actuator status data. Material characteristics data reflects the physical properties of the lithium ore itself, including particle size distribution, moisture content, bulk density, and hardness. Particle size distribution data is collected by a laser particle size analyzer installed at the hopper inlet. This instrument uses the principle of laser scattering to analyze the proportion of ore particles within different particle size ranges. For example, it can distinguish the distribution of particles in different size ranges such as 0-5mm, 5-10mm, and 10-20mm. Moisture content data is obtained by a microwave moisture meter, which monitors the moisture content of the ore in real time by measuring the difference in microwave propagation speed in ores with different moisture contents. Bulk density data is collected using a weighing sensor in conjunction with a volume measurement device, first measuring the weight of a certain volume of ore and then calculating the mass per unit volume. Hardness data is collected by an impact hardness tester installed above the conveyor belt, which determines the ore hardness based on the feedback force when the probe contacts the ore surface.

[0031] Equipment operation data covers the real-time operating status of each core device in the conveying system, including motor speed, output torque, winding temperature; conveyor belt speed, tension, and deviation; hopper vibration frequency and amplitude; and gearbox oil temperature and level. This data is collected by encoders installed on the motor shaft, tension sensors on both sides of the conveyor belt, and vibration sensors on the hopper's outer wall, providing a real-time view of the equipment's operating status.

[0032] Energy consumption monitoring data is used to record the system's energy consumption, including the motor's active power, reactive power, and power factor, the energy consumption of the conveyor belt drive system, the control system's standby power consumption, and the total power consumption of the entire conveyor system. This data is collected by smart meters, power sensors, and other devices within the motor control cabinet, and recorded and stored at regular intervals.

[0033] Environmental sensing data primarily includes environmental parameters surrounding the hopper, such as ambient temperature, relative humidity, dust concentration, and air pressure. Temperature and humidity are collected using temperature and humidity sensors, dust concentration is monitored using a laser dust sensor, and air pressure data is acquired using a pressure sensor. This data is used to analyze the potential impact of environmental factors on equipment operation and ore characteristics.

[0034] The actuator status data involves the status of components directly involved in the regulating action, including the real-time opening of the gate, the current and temperature of the stepper motor that drives the gate, the pressure and flow of the hydraulic device that performs the regulating action, and the fault alarm signals of each actuator, etc., which are collected through position sensors, current sensors, pressure sensors, etc. installed on the actuator.

[0035] When constructing the hopper conveying status database, the collected multidimensional operating data is first preprocessed. For outliers, statistical analysis is performed to determine the normal fluctuation range of each data point. Data that exceeds this range by a certain multiple is identified as an outlier and removed. For example, when the motor power at a certain moment is significantly higher than the power value under the same operating conditions during the same historical period, it is considered an outlier. For missing values, if the amount of missing data is small, linear interpolation is performed using valid data from adjacent moments. If data is continuously missing within a certain time period and the data difference between adjacent moments is large, the data under the same operating conditions before and after the time period is used to fill the gap to ensure data continuity.

[0036] After preprocessing, the data is dimensionalized. Because different types of data have different units and magnitudes—for example, motor power is measured in kW and temperature is measured in °C—directly using these values ​​can affect the accuracy of subsequent analysis. Therefore, all data must be converted to a unified numerical range. Specifically, a linear transformation is used to convert each data value to a dimensionless value between 0 and 1. This conversion preserves the relative size of the data, allowing different types of data to be compared on the same scale.

[0037] At the same time, the pre-processed operating data is time-series aligned. Because different sensors may have different sampling frequencies, resulting in deviations in the timestamps of different data at the same time, it is necessary to use the system's unified clock as a benchmark and match all data by timestamp to ensure that the various types of data collected at the same time can accurately correspond. For example, the data collected by the particle size sensor every 100ms and the motor power data collected every 50ms are aligned according to millisecond timestamps to form a unified data sequence in the time dimension and generate standardized feature data.

[0038] Finally, all standardized feature data is structured and stored according to the time and space dimensions. In the time dimension, millisecond-level timestamps are used as indexes and arranged in chronological order to form a continuous time series data sequence, which can trace the system operation status at any time. In the spatial dimension, sensors are classified according to their installation location and monitoring objects. For example, sensor data at the hopper entrance, sensor data in the middle of the conveyor belt, and sensor data near the motor are classified and stored separately. Through this structured storage method, all standardized feature data are integrated into a complete hopper conveying status database. This database not only contains historical operation data, but also receives and updates newly collected data in real time, providing comprehensive and accurate data support for subsequent dynamic feature extraction and control decisions.

[0039] Example 2: See Figure 3 When extracting dynamic features from the hopper conveying status database, a sliding window algorithm is used to process the time series data within it. The sliding window size is set to a fixed time period. This time period is selected based on the system's operational stability and the frequency of data changes, ensuring that each window contains sufficient information to reflect the operational characteristics within that period while also preventing excessive window size from obscuring dynamic data changes. The sliding step size is determined based on the response speed requirements of real-time control to ensure timely capture of system state changes. In practice, the chronologically arranged time series data in the hopper conveying status database is intercepted according to the set window size and sliding step size, with each window corresponding to a specific time period. For each window of time series data, local feature vectors are extracted that reflect the operational status within that period. These local feature vectors contain statistics from multiple dimensions, such as the average motor power, the maximum and minimum conveyor belt speed, the fluctuation range of energy consumption data, and the concentration trend of the material particle size distribution. These statistics are used to characterize the local operational characteristics of the system within that period.

[0040] After obtaining the local eigenvectors for each time period, a real-time control parameter matrix is ​​constructed based on the preset control parameter thresholds. The control parameter thresholds are determined based on the system's design parameters, the equipment's performance indicators, and actual operating experience, covering various parameter ranges for motor operation, the reasonable operating speed range of the conveyor belt, the effective adjustment range of the gate opening, and so on. The local eigenvectors for each time period are compared and matched with these preset control parameter thresholds to screen out parameters that meet the threshold requirements and are of practical significance to system control. These parameters are then arranged in chronological order and by parameter type to form a real-time control parameter matrix. The rows of this matrix represent different time periods, and the columns represent different control parameters. Each element in the matrix is ​​the specific value of the control parameter within the corresponding time period, thus clearly displaying the control parameter status of the system at different times.

[0041] When establishing an association rule diagram for each parameter in the real-time control parameter matrix, target control parameters are selected from the matrix. Priority is given to operating parameters directly related to energy consumption indicators, such as motor operating power, conveyor belt operating speed, and gate opening degree, as these parameters have a direct and significant impact on system energy consumption. By analyzing the relationship between these parameters and other parameters, key links in energy-saving control can be more effectively identified.

[0042] After determining the target control parameter, all associated parameters related to the target control parameter are obtained. This requires traversing the other parameters in the real-time control parameter matrix and analyzing the time series trends and correlations between the parameters to identify parameters that may interact or be associated with the target control parameter. For example, when motor operating power is used as the target control parameter, related associated parameters may include conveyor belt operating speed, material bulk density, ambient temperature, etc., because changes in these parameters may lead to changes in motor operating power.

[0043] The Pearson correlation coefficient between the target control parameter and each associated parameter is calculated to determine the strength of the parameter association. The correlation coefficient between the target control parameter and each associated parameter is calculated by statistically analyzing the numerical sequences. The value range of this coefficient is between -1 and 1. The closer the absolute value is to 1, the stronger the linear correlation between the two parameters. A positive coefficient indicates a positive correlation, that is, when one parameter increases, the other parameter also increases. A negative coefficient indicates a negative correlation, that is, when one parameter increases, the other parameter decreases.

[0044] Based on the calculated parameter association strength and the preset association threshold, an association rule subgraph for the target control parameter is constructed. The association threshold is set based on the requirements for parameter association in actual applications. When the absolute value of the correlation coefficient between two parameters is greater than the threshold, a significant association relationship is considered to exist between them. These two parameters are treated as nodes and connected by an edge. The edge weight is represented by the absolute value of the correlation coefficient. This is how the association rule subgraph for a single target control parameter is constructed.

[0045] The association rule subgraphs for all target control parameters are integrated to form the association rule graph for the real-time control parameter matrix. During the integration process, identical nodes and edges are merged, and for duplicate edges, the average weight is taken as the final weight. This creates a complete graph structure that reflects the association relationships between all parameters in the real-time control parameter matrix.

[0046] Example 3: When extracting the key influencing factors of each control parameter from the association rule graph, the random forest algorithm is used to calculate the importance score of each feature. Specifically, all parameters in the association rule graph are used as input features. These parameters include motor operating power, conveyor belt speed, gate opening, material particle size distribution, ambient temperature, etc., and the comprehensive performance of energy consumption and efficiency of the conveying system is used as the output label. In the process of training the random forest model, the contribution of each input feature in the decision-making process is quantitatively evaluated through integrated learning of multiple decision trees, and finally the importance score of each feature is output. The score reflects the importance of the corresponding feature in predicting the output label. The higher the score, the more significant the impact of the feature on the system operation status.

[0047] A preset threshold is set to screen key influencing factors. This threshold is determined based on the average contribution of each feature to control decisions in historical regulation data. Within this historical regulation data, the contribution ratio of each feature to the decision outcome under different regulation scenarios is calculated, and the average of these ratios is calculated and used as the preset threshold. The importance score of each feature output by the random forest algorithm is compared with the preset threshold. Features with scores above the threshold are selected as key influencing factors. Together, these factors form the core set of parameters that influence system control decisions.

[0048] When determining the weight coefficient for each key influencing factor, a normalization process is performed based on the feature importance score. The importance score of each key influencing factor is divided by the sum of the importance scores of all key influencing factors. The result is the weight coefficient of that factor. The size of the weight coefficient directly reflects the relative importance of the corresponding key influencing factor in the control decision. Factors with higher scores have larger weight coefficients and occupy a more dominant position in the subsequent optimization model.

[0049] A multi-objective optimization model was established based on a preset adjustment objective function. The adjustment objective function includes two core goals: minimizing unit material energy consumption and maximizing conveying capacity. Unit material energy consumption is measured as the electrical energy consumed per unit mass of lithium ore transported, while conveying capacity is measured as the mass of lithium ore passing through the conveyor belt per unit time. The model's constraints encompass the physical limitations of the equipment, such as the maximum motor speed limiting the maximum operating speed of the conveyor belt and the maximum gate opening limiting the maximum drop rate of the material. It also includes process requirements, such as a limit on the breakage rate during material transportation.

[0050] The multi-objective optimization model is solved using the particle swarm optimization algorithm. A certain number of particles are initialized, with each particle representing a set of control parameter combinations, including specific values ​​for motor speed, gate opening, and conveyor belt tension. The particles iteratively search for optimal solution within the solution space, adjusting their flight speed and direction based on their own historical optimal position and the global optimal position of the entire swarm. During this iteration, each particle's fitness is calculated based on the adjusted objective function. This fitness value comprehensively considers the performance of unit material energy consumption and conveying volume. A higher fitness value indicates a more optimal control parameter combination represented by the particle.

[0051] In order to balance the two objectives, the objective weight coefficient is introduced, and the calculation formula is as follows:

[0052] in, is the comprehensive fitness value of the particle, is the weight coefficient of the unit material energy consumption target, is the normalized value of unit material energy consumption, is the normalized value of the transport volume.

[0053] During the iteration process, the particle positions and velocities are continuously updated until a pre-defined termination condition is met, such as reaching the maximum number of iterations or the optimal fitness value of the particle swarm no longer changes significantly. Ultimately, the control parameter combination represented by the optimal particle obtained through iteration is used as the model output to generate an intelligent regulation decision model. This model receives real-time input of system operating data and, through internal computational logic, outputs corresponding regulation strategies to achieve dynamic control of the lithium ore hopper conveying system.

[0054] Example 4: See Figure 4When calculating the energy consumption impact coefficient of each operating parameter, it is first necessary to obtain two types of data, namely the historical energy consumption baseline value of each operating parameter and the current energy consumption actual value. The historical energy consumption baseline value comes from the operating data under the same operating conditions in the past period of time. Specifically, the operating time periods in the past 30 days that are similar to the current material properties and environmental conditions are selected, and the energy consumption data in these time periods are averaged to obtain the value. For example, when the particle size distribution of the currently transported lithium ore is concentrated in 5-10mm and the ambient temperature is 25°C, the energy consumption records under the same particle size range and temperature conditions will be screened out from the historical data, and the average value of these records will be calculated as the historical energy consumption baseline value. The current energy consumption actual value is collected in real time through sensors installed on the equipment, including the real-time power of the motor, the energy consumption of the conveyor belt drive system, etc. The collection frequency is consistent with the sampling frequency of the data acquisition module to ensure the real-time nature of the data.

[0055] After obtaining the historical energy consumption baseline value and the current energy consumption actual value, the deviation rate between the two is calculated. The deviation rate reflects the degree of deviation between the current energy consumption and the historical benchmark energy consumption. At the same time, the contribution of each feature is extracted from the key energy consumption feature vector. These contributions are obtained by analyzing the proportion of the impact of different parameters on the total energy consumption in historical operation. For example, the characteristic contribution of motor power may be higher than the characteristic contribution of ambient temperature, because motor operation directly consumes electrical energy, while ambient temperature only indirectly affects energy consumption by affecting the heat dissipation of the equipment. The deviation rate and the characteristic contribution are weighted multiplied, and the result obtained is the energy consumption impact coefficient of each operating parameter. When the coefficient is positive, it means that the current energy consumption of the parameter is higher than the historical benchmark level; when the coefficient is negative, it means that the current energy consumption is lower than the historical benchmark level, and the absolute value of the coefficient reflects the strength of the impact of the parameter on the overall energy consumption.

[0056] The process of selecting the optimal energy-saving control sequence begins by inputting the energy consumption impact coefficients of various operating parameters into the intelligent regulation decision-making model. Based on these coefficients and combined with its internal algorithmic logic, the model generates multiple candidate control sequences. Each candidate control sequence contains multiple specific operating parameter settings. For example, one sequence might set the motor speed to 1500 rpm, the gate opening to 30%, and the conveyor speed to 2 m / s. Another sequence might set the motor speed to 1400 rpm, the gate opening to 25%, and the conveyor speed to 1.8 m / s. Different parameter combinations correspond to different operating states.

[0057] A system dynamics model is used to simulate the energy consumption of each candidate control sequence. This model considers multiple factors that influence energy consumption, including the motor's efficiency curve at different speeds—that is, the motor's efficiency at converting electricity to electricity at a specific speed; conveyor belt friction loss. As the conveyor belt speed changes, the friction between it and the support rollers changes, which in turn affects energy consumption; and the effect of gate opening on the material's falling speed. A larger opening increases the material's falling speed, which can cause changes in motor load and, consequently, energy consumption. By comprehensively simulating these factors, the energy consumption per unit time for each candidate control sequence is calculated.

[0058] To evaluate conveying efficiency, a discrete event simulation method was used. The moment lithium ore enters the hopper was used as the event trigger. The material's movement along the conveyor belt was recorded, including the time it takes for the material to fall from the hopper to the conveyor belt, the time it spends on the conveyor belt, and the time it reaches the destination. By summarizing these time data, the average material conveying speed was calculated: the ratio of total conveying distance to total conveying time. The pass rate was also calculated by measuring the ratio of the amount of material successfully passing through the conveyor belt to the total amount of material entering the hopper. This pass rate reflects the conveying system's material handling capacity and stability. A low pass rate may indicate problems such as material blockage or material drop during transportation.

[0059] After completing the energy consumption simulation calculation and transmission efficiency evaluation, a comprehensive performance score is given to each group of candidate control sequences. During the scoring process, the energy consumption index and efficiency index are weighted according to a certain ratio. For example, the energy consumption index accounts for 60% and the efficiency index accounts for 40%. The specific ratio is determined according to the priority requirements for energy saving and efficiency in actual production. In the energy consumption index, the lower the energy consumption per unit time, the higher the score; in the efficiency index, the faster the average speed and the higher the pass rate, the higher the score. The two scores are added together to obtain the comprehensive performance score of each group of candidate control sequences. Finally, the candidate control sequence with the highest score is selected as the optimal energy-saving control sequence. This sequence can minimize energy consumption while meeting the transmission efficiency and achieve optimized operation of the system.

[0060] Example 5: The actuator consists of a variable frequency speed regulating motor, an electric regulating gate and a pressure sensor feedback unit. The variable frequency speed regulating motor serves as the power source of the conveyor belt. It adjusts the output speed by changing the frequency of the input power supply, thereby controlling the running speed of the conveyor belt. It has an integrated speed feedback device that can monitor its actual speed in real time and feed this information back to the control system. The electric regulating gate is installed at the outlet of the lithium ore hopper and is driven by a stepper motor. The degree of opening of the gate is adjusted by precisely controlling the rotation angle of the stepper motor, thereby changing the speed and flow of the falling material. A position sensor is installed on the gate to detect the actual opening of the gate in real time to ensure the adjustment accuracy. The pressure sensor feedback unit includes a plurality of pressure sensors evenly distributed under the conveyor belt. It can sense the weight and distribution of the material on the conveyor belt, convert the pressure signal into an electrical signal and transmit it to the control system, providing real-time material load information for adjustment decisions.

[0061] When driving the actuator to complete the adaptive adjustment of the hopper conveying parameters according to the optimal energy-saving control sequence, it is first necessary to convert the various parameter values ​​in the optimal energy-saving control sequence into control instructions that the actuator can recognize. The optimal energy-saving control sequence contains target parameters such as conveyor belt speed, gate opening, and motor frequency. These parameters need to be converted into signal forms that match the actuator. For example, the target value of the conveyor belt speed needs to be converted into a frequency instruction for the variable frequency speed regulation motor, because the output speed of the motor is proportional to the input frequency. A specific frequency corresponds to a specific speed, which in turn determines the running speed of the conveyor belt. The target value of the gate opening needs to be converted into a pulse number instruction for the stepper motor that electrically adjusts the gate. Every time the stepper motor receives a certain number of pulses, it will rotate the corresponding angle, thereby driving the gate to reach the target opening.

[0062] After a control instruction is generated, it is sent to the corresponding execution unit via the industrial bus. The industrial bus utilizes a high-performance communication protocol, enabling rapid and stable transmission of control instructions, ensuring that each execution unit can promptly receive and respond to the instruction. After the variable frequency speed motor receives the frequency instruction, its internal frequency conversion module adjusts the frequency of the output power supply, gradually bringing the motor's speed closer to the target speed, thereby driving the conveyor belt speed to the value set by the optimal energy-saving control sequence. During the adjustment process, the motor's speed feedback device continuously transmits actual speed information back to the control system, forming a closed-loop control loop to minimize the deviation between the actual and target speeds.

[0063] After receiving a pulse count command, the electric gate adjusts the gate's opening position. The stepper motor begins rotating according to the command, driving the gate blades through a mechanical transmission mechanism to adjust the gate's opening degree. A position sensor on the gate monitors the actual gate opening in real time and feeds this information back to the control system. The control system compares the actual opening with the target opening and, if any deviation exists, issues a correction command until the gate reaches the target opening. This process ensures accurate gate opening adjustment and prevents over- or under-feeding of materials due to improper opening, which could impact system energy consumption and efficiency.

[0064] The pressure sensor feedback unit operates continuously throughout the entire adjustment process, collecting real-time data on the material weight on the conveyor belt and converting this data into electrical signals for transmission to the control system. The control system analyzes this data to understand the current material load and determine whether the parameters in the optimal energy-saving control sequence are appropriate for the actual material volume. If the material load undergoes significant changes, such as a sudden increase or decrease, the control system reassesses the optimal energy-saving control sequence based on the feedback from the pressure sensor. It then generates new control instructions and sends them to the actuators, re-adjusting parameters such as conveyor belt speed and gate opening, achieving dynamic adaptive adjustment of conveying parameters.

[0065] Through the above process, the execution and adjustment module can drive the various actuators to work together according to the optimal energy-saving control sequence, so that the various parameters of the lithium ore hopper conveying system are always in a state that matches the material characteristics and load conditions, thereby achieving adaptive adjustment.

[0066] It should be noted that, in this document, relational terms such as first and second, etc., are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the terms "comprises," "comprising," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that includes a list of elements includes not only those elements but also other elements not explicitly listed, or elements inherent to such process, method, article, or apparatus.

[0067] While 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 these embodiments without departing from the principles and spirit of the invention, and that the scope of the invention is defined by the appended claims and their equivalents.

Claims

1. The lithium ore hopper energy-saving conveying system based on intelligent adaptive control is characterized by: The delivery system comprises: A data acquisition module is used to collect multi-dimensional operating data of the lithium ore hopper through a distributed sensor network, standardize the multi-dimensional operating data, and build a hopper conveying status database; An adaptive control module is used to extract dynamic features from the hopper conveying state database, generate a real-time control parameter matrix, establish an association rule graph for each parameter in the real-time control parameter matrix, and construct an intelligent adjustment decision model based on the association rule graph; An energy-saving optimization module, configured to extract key energy consumption feature vectors from the real-time control parameter matrix, calculate the energy consumption impact coefficient of each operating parameter based on the key energy consumption feature vectors, and select the optimal energy-saving control sequence in combination with the intelligent adjustment decision model; The execution adjustment module is used to drive the actuator to complete the adaptive adjustment of the hopper conveying parameters according to the optimal energy-saving control sequence.

2. The lithium ore hopper energy-saving conveying system based on intelligent adaptive control according to claim 1 is characterized in that: The multi-dimensional operation data of the lithium ore hopper includes material property data, equipment operation data, energy consumption monitoring data, environmental perception data and actuator status data; the material property data includes particle size distribution data, moisture content data, bulk density data and hardness data of the lithium ore.

3. The lithium ore hopper energy-saving conveying system based on intelligent adaptive control according to claim 1 is characterized in that: Standardizing the multi-dimensional operation data and constructing a hopper conveying status database specifically includes: Eliminating outliers and filling missing values ​​on the multidimensional operating data to obtain preprocessed operating data; Perform dimension unification and time-series alignment on the pre-processed operating data to generate standardized feature data; All standardized feature data are structured and stored according to the time dimension and space dimension to build a hopper conveying status database.

4. The lithium ore hopper energy-saving conveying system based on intelligent adaptive control according to claim 1 is characterized in that: Extracting dynamic features from the hopper conveying state database to generate a real-time control parameter matrix specifically includes: A sliding window algorithm is used to extract segmented features from the time series data in the hopper conveying state database to obtain local feature vectors for each time period; A real-time control parameter matrix is ​​constructed based on the local feature vectors of each time period and the preset control parameter thresholds.

5. The lithium ore hopper energy-saving conveying system based on intelligent adaptive control according to claim 1 is characterized in that: Establishing an association rule diagram for each parameter in the real-time control parameter matrix specifically includes: Selecting a target control parameter from the real-time control parameter matrix and acquiring all associated parameters related to the target control parameter; Calculate the Pearson correlation coefficient between the target control parameter and each associated parameter to determine the parameter association strength; An association rule subgraph of the target control parameters is constructed according to the parameter association strength and the preset association threshold, and an association rule graph of the real-time control parameter matrix is ​​obtained by integrating the association rule subgraphs of all target control parameters; When selecting target control parameters in the real-time control parameter matrix, operating parameters directly related to energy consumption indicators are preferably selected as target control parameters. The operating parameters directly related to energy consumption indicators include motor operating power, conveyor belt operating speed and gate opening degree.

6. The lithium ore hopper energy-saving conveying system based on intelligent adaptive control according to claim 1 is characterized in that: Constructing an intelligent adjustment decision model based on the association rule graph specifically includes: Extracting key influencing factors of each control parameter from the association rule graph and determining weight coefficients of each key influencing factor; Establish a multi-objective optimization model based on the weight coefficients of key influencing factors and the preset adjustment objective function; Solving the multi-objective optimization model by using a particle swarm optimization algorithm to generate an intelligent adjustment decision model; When extracting the key influencing factors of each control parameter from the association rule graph, the importance score of each feature is calculated by the random forest algorithm, and features with scores higher than a preset threshold are selected as key influencing factors. The preset threshold is determined based on the average feature contribution of historical adjustment data.

7. The lithium ore hopper energy-saving conveying system based on intelligent adaptive control according to claim 1 is characterized in that: Calculating the energy consumption impact coefficient of each operating parameter based on the key energy consumption characteristic vector specifically includes: Obtain the historical energy consumption benchmark value and current energy consumption measured value of each operating parameter; Calculate the deviation rate between the current energy consumption measured value and the historical energy consumption benchmark value, combined with the characteristic contribution of the key energy consumption feature vector; The energy consumption impact coefficient of each operating parameter is determined by the weighted product of the deviation rate and the characteristic contribution.

8. The lithium ore hopper energy-saving conveying system based on intelligent adaptive control according to claim 1 is characterized in that: The optimal energy-saving control sequence selected by combining the intelligent adjustment decision model specifically includes: Input the energy consumption impact coefficient of each operating parameter into the intelligent regulation decision model to generate multiple sets of candidate control sequences; Perform energy consumption simulation and transmission efficiency evaluation on each set of candidate control sequences to obtain a comprehensive performance score; The candidate control sequence with the highest comprehensive performance score is selected as the optimal energy-saving control sequence; The energy consumption simulation calculation and transportation efficiency evaluation of each set of candidate control sequences specifically include: simulating the energy consumption per unit time under the candidate control sequence through the system dynamics model, and using the discrete event simulation method to evaluate the average speed and throughput rate of material transportation.

9. The lithium ore hopper energy-saving conveying system based on intelligent adaptive control according to claim 1, characterized in that: Driving the actuator to complete the adaptive adjustment of the hopper conveying parameters according to the optimal energy-saving control sequence is to convert the optimal energy-saving control sequence into a control instruction of the actuator, and send it to the corresponding execution unit through the industrial bus to achieve coordinated adjustment of the conveying speed, gate opening and motor frequency.

10. The lithium ore hopper energy-saving conveying system based on intelligent adaptive control according to claim 9, characterized in that: The actuator includes a variable frequency speed regulating motor, an electric regulating gate and a pressure sensor feedback unit.

Images