Coal resource scheduling method and system

Abstract

The application relates to the technical field of coal resource scheduling, in particular to a coal resource scheduling method and system. The method comprises the following steps: acquiring a running state signal, a theoretical rotating speed of a driving motor, a theoretical belt speed converted according to the theoretical rotating speed, an instantaneous unit length material weight output by a weighing device and economic parameters; calculating an actual running speed of a conveying belt according to the running state signal; calculating a conveying capacity deviation based on the actual running speed and the instantaneous unit length material weight; calculating extra energy consumption in a monitoring period; calculating an environmental protection agent consumption amount, and calculating a downstream compensation cost in combination with the economic parameters; generating an economic benefit and energy consumption loss report, and outputting a maintenance suggestion or a scheduling correction instruction. The method aims to solve the problem that, in the existing coal resource scheduling method, the conveying capacity deviation cannot be identified in time, the instantaneous conveying amount calculation error is caused, and finally the scheduling system optimization effect and clean and efficient utilization of coal are affected.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of coal resource scheduling, in particular to a coal resource scheduling method and system. BACKGROUND

[0002] In large-scale industrial production, coal resource scheduling relies on the premise of stable and predictable maximum conveying capacity of key equipment (especially the main conveying belt). The scheduling and optimization model plans the transportation time, mixing ratio and material rhythm by combining the theoretical belt speed converted by the driving motor speed with the weighing data, assuming that the actual belt speed = the theoretical speed.

[0003] However, long-term heavy load operation causes uneven wear of the driving drum rubber layer, the friction between the drum and the belt decreases, forming slight and continuous slipping, which causes the actual speed of the belt to be lower than the theoretical speed. The existing monitoring relies more on motor current and speed, there is no independent belt surface speed measurement, and there is also a lack of identification path for small and continuous deviations, so this sub-health state is difficult to expose in time. As a result, there is a chain deviation: the scheduling system converts the instantaneous conveying capacity based on the high theoretical speed, resulting in systematic overestimation of the conveying data; the optimization program allocates slightly over-capacity planned tasks accordingly; the downstream mixed bin / boiler bin is short of material, and the operator is forced to exceed the optimization strategy to make empirical and early shipment, disrupting the overall rhythm, weakening the clean and efficient goal, and potentially increasing energy consumption, fuel and environmental protection agent costs.

[0004] In view of the above problems, the existing technology needs to be improved. SUMMARY

[0005] The present application discloses a coal resource scheduling method and system, aiming to solve the problem that in the existing coal resource scheduling method, the belt slipping caused by uneven wear of the driving drum rubber layer of the conveying belt cannot be identified in time, resulting in calculation error of instantaneous conveying capacity, and ultimately affecting the optimization effect of the scheduling system and the clean and efficient use of coal.

[0006] The technical scheme of the present application is as follows:

[0007] In a first aspect, the present application discloses a coal resource scheduling method, comprising the following steps:

[0008] Obtaining the running state signal of the driving drum area of the conveying belt, the theoretical speed of the driving motor, and the theoretical belt speed converted therefrom, the instantaneous unit length material weight output by the weighing device, and the economic parameters;

[0009] According to the running state signal, the belt slipping related features are obtained, the belt slipping related features are combined with the theoretical belt speed and the instantaneous unit length material weight, and the actual running speed of the conveying belt is calculated according to the pre-labeled mapping relationship.

[0010] The corrected instantaneous material conveying amount is calculated based on the actual running speed and the instantaneous unit length material weight, and the relative difference between the theoretical belt speed and the actual running speed is used to represent the conveying capacity deviation;

[0011] The real-time power of the driving motor is monitored, and the additional energy consumption in the monitoring period is calculated according to the reference power established under the no-slip working condition;

[0012] The running state and energy consumption of the downstream associated equipment, the fuel consumption of the downstream coal-using equipment, and the environmental protection agent consumption of the flue gas treatment device are monitored, and the downstream compensation cost caused by the conveying capacity deviation is calculated in combination with the economic parameters;

[0013] The additional energy consumption and the downstream compensation cost are summarized and the cumulative economic loss is calculated to generate an economic benefit and energy consumption loss report;

[0014] When the cumulative economic loss and / or the conveying capacity deviation reaches a preset threshold, a maintenance suggestion or a dispatching correction instruction is output.

[0015] Further, according to the running state signal, a belt slip related feature is obtained, and the belt slip related feature is combined with the theoretical belt speed and the instantaneous unit length material weight to calculate the actual running speed of the conveying belt according to a pre-labeled mapping relationship, including:

[0016] The running state signal includes a multi-channel acoustic data stream collected by a multi-channel acoustic sensor array arranged in the driving drum area of the conveying belt, the multi-channel acoustic data stream is spatially positioned, and acoustic characteristic calibration is performed to obtain the spatial direction of the slip sound source and at least one noise source and the calibration parameters for spatial filtering;

[0017] Based on the spatial direction and the calibration parameters, the multi-channel acoustic data stream is spatially filtered to enhance the signal from the slip sound source direction and attenuate the signal from the noise source direction, and a focused signal is obtained;

[0018] The belt slip related feature is extracted from the focused signal;

[0019] The belt slip related feature is combined with the theoretical belt speed and the instantaneous unit length material weight to calculate the actual running speed of the conveying belt according to the mapping relationship.

[0020] Further, the multi-channel acoustic data stream is spatially positioned, and acoustic characteristic calibration is performed to obtain the spatial direction of the slip sound source and at least one noise source and the calibration parameters for spatial filtering, including:

[0021] According to the multi-channel acoustic data stream combined with the running state of the conveyor belt, the start-stop state of the dust suppression system, and the running state information of the noise-related equipment, specific time periods with only skid sound sources or only noise sources are identified;

[0022] In a specific time period, the current spatial direction of the skid sound source and at least one noise source relative to the multi-channel acoustic sensor array is determined using a time difference of arrival and / or a generalized cross-correlation algorithm;

[0023] In a specific time period, the spectral characteristics, energy distribution, and directivity characteristics of the skid sound source and the noise source are analyzed to obtain their current calibration parameters;

[0024] When a new noise source is detected or the spatial direction of an existing noise source changes, spatial positioning and acoustic characteristic calibration of the new or changed noise source are triggered to update the spatial direction and calibration parameters;

[0025] According to the real-time determined spatial direction and calibration parameters, the parameters of the spatial filtering process, including beamforming weights and / or null directions, are updated to maintain the signal-to-noise ratio of the focused signal and continuously enhance the signal from the direction of the skid sound source and attenuate the signal from the direction of the noise source.

[0026] Further, according to the real-time determined spatial direction and calibration parameters, the parameters of the spatial filtering process, including beamforming weights and / or null directions, are updated, including:

[0027] Real-time monitoring of the movement trajectory and transient acoustic characteristics of the noise source, the movement trajectory being determined by the change of the spatial direction of the noise source relative to the multi-channel acoustic sensor array over time, and the transient acoustic characteristics being determined by the time variation of the spectrum and energy distribution of the multi-channel acoustic data stream;

[0028] Based on the movement trajectory, the spatial direction of the noise source in the next time period is predicted;

[0029] Based on the transient acoustic characteristics, the calibration parameters of the noise source in the next time period are predicted;

[0030] Based on the predicted spatial direction and calibration parameters, the parameters for the spatial filtering process in the next time period are pre-calculated in the current time period, including beamforming weights and / or null directions.

[0031] Further, when the cumulative economic loss and / or the deviation of the conveying capacity reaches a preset threshold, a maintenance recommendation or a dispatch correction instruction is output, including:

[0032] Based on the economic benefit and energy consumption loss report, the growth rate of the cumulative economic loss and the trend index of the deviation of the conveying capacity are calculated;

[0033] A cost prediction model is established in combination with economic parameters to predict the changes of the unit prices of electric energy, fuel and environmental protection agent in the pre-design planning period, and a cost prediction sequence is obtained;

[0034] In the preset multiple scenarios, the total cost in the planning period is estimated based on the cost prediction sequence and the growth rate and trend indicators, and the total cost in the planning period includes the predicted additional energy consumption and downstream compensation cost, and the maintenance cost and downtime loss cost;

[0035] A candidate maintenance time point and a dynamic adjustment amount of the preset threshold value are determined to minimize the difference of the total cost in the planning period, and a maintenance condition is defined, which at least includes: in any candidate maintenance time point, the difference of the total cost in the planning period of the multiple scenarios is not less than the preset maintenance triggering threshold, and / or the growth rate of the cumulative economic loss and the trend indicator of the deviation of the conveying capacity meet the preset joint threshold;

[0036] When the maintenance condition is met, a maintenance suggestion is generated;

[0037] When the maintenance condition is not met, a scheduling correction instruction is formed, which at least includes: adjusting the operation time and start-stop sequence of the downstream associated equipment, adjusting the fuel consumption plan of the downstream coal-using equipment, and adjusting the environmental protection agent addition strategy of the flue gas treatment device.

[0038] Further, the moving trajectory and transient acoustic characteristics of the noise source are monitored in real time, including:

[0039] In the initialization stage or when the noise source changes are detected, spatial positioning and acoustic characteristic calibration are performed for each noise source respectively to obtain the initial spatial direction and calibration parameters thereof;

[0040] A multi-target sound source separation algorithm is applied, and the signal components corresponding to each noise source in the multi-channel acoustic data stream are separated into independent acoustic signals corresponding to each noise source in combination with the initial spatial direction and calibration parameters of each noise source;

[0041] The transient acoustic characteristics of each separated independent acoustic signal are extracted through time-frequency analysis;

[0042] Using each independent acoustic signal, the current spatial direction of each noise source relative to the multi-channel acoustic sensor array is determined and continuously updated by using the time difference of arrival and / or generalized cross-correlation algorithm, so as to construct the moving trajectory of each noise source.

[0043] Further, a multi-target sound source separation algorithm is applied, and the signal components corresponding to each noise source in the multi-channel acoustic data stream are separated into independent acoustic signals corresponding to each noise source in combination with the initial spatial direction and calibration parameters of each noise source, including:

[0044] Combine the multi-channel acoustic data stream, the running state of the conveyor belt, the corrected instantaneous material conveying capacity, and the real-time power of the driving motor to determine whether a transient change occurs in the acoustic scene;

[0045] When it is determined that a transient change occurs in the acoustic scene, perform sound source activity detection on the multi-channel acoustic data stream to identify a potential active sound source region;

[0046] Compare the acoustic characteristics of the potential active sound source region with a preset slip sound source fingerprint library and an environmental background noise fingerprint library to obtain a category determination and / or a similarity index;

[0047] For a sound source region with a similarity to the slip sound source fingerprint library or the environmental background noise fingerprint library greater than a preset similarity threshold, assign a lower noise source confidence weight and / or exclude it from the noise source separation target in the multi-target sound source separation algorithm according to a preset weight rule;

[0048] For a sound source region with a similarity to the slip sound source fingerprint library or the environmental background noise fingerprint library less than a preset similarity threshold and a similarity to at least one noise source fingerprint in the preset noise source fingerprint library greater than a preset noise threshold, add it as a new noise source to the noise source separation target, and assign a higher noise source confidence weight in the multi-target sound source separation algorithm according to a preset weight rule;

[0049] During the execution of the multi-target sound source separation algorithm, dynamically adjust the separation strategy according to the noise source confidence weight;

[0050] Cross-verify each independent acoustic signal obtained by separation, which includes comparing with historical noise source moving tracks, acoustic fingerprints, and / or running state information of noise-associated equipment to confirm whether it is a real noise source.

[0051] Further, cross-verify each independent acoustic signal obtained by separation, which includes comparing with historical noise source moving tracks, acoustic fingerprints, and / or running state information of noise-associated equipment to confirm whether it is a real noise source, including:

[0052] Obtain the running state information of the noise-associated equipment, including the start-stop state and power consumption of the noise-associated equipment;

[0053] According to a preset multi-dimensional verification rule set, cross-verify the transient acoustic characteristics, moving tracks, and running state information of the noise-associated equipment to obtain a cross-verification determination result;

[0054] According to the cross-verification determination result, confirm whether each independent acoustic signal corresponds to a real noise source, and correct false positives or false negatives for situations that do not meet the determination conditions.

[0055] Further, according to the cross-validation determination result, it is determined whether each independent acoustic signal corresponds to a real noise source, and false positives or false negatives in cases not meeting the determination condition are corrected, including:

[0056] When the cross-validation determination result indicates that the transient acoustic characteristics of the independent acoustic signal match the preset noise source fingerprint library, but the operation state information of the noise-related device is inconsistent or abnormal, the corresponding independent acoustic signal is marked as a to-be-confirmed state;

[0057] For the to-be-confirmed state, an abnormal mode analysis is performed, and the historical correlation between the transient acoustic characteristics, the moving track and the operation state information of the noise-related device is compared within a preset time window to obtain a historical correlation analysis result;

[0058] When the historical correlation analysis result shows that the independent acoustic signal appears at the time when the noise-related device is abnormal or not started, and matches the transient acoustic characteristics collected at the end time of the preset time window, the independent acoustic signal is confirmed as a real noise source;

[0059] When the historical correlation analysis result shows that the independent acoustic signal does not appear when the noise-related device is abnormal or not started, and / or its transient acoustic characteristics do not match the transient acoustic characteristics collected at the end time of the preset time window, the independent acoustic signal is marked as a non-real noise source;

[0060] When the operation state information of the noise-related device returns to normal, the independent acoustic signal in the to-be-confirmed state is cross-validated according to the multi-dimensional verification rule set, and the confirmation conclusion of whether it is a real noise source is updated.

[0061] In a second aspect, the application also discloses a coal resource scheduling system, comprising:

[0062] A signal acquisition module is configured to acquire a running state signal of a driving drum area of a conveyor belt, a theoretical speed of a driving motor and a theoretical belt speed converted therefrom, an instantaneous unit length material weight output by a weighing device, and economic parameters;

[0063] A speed calculation module is configured to obtain a belt slip related feature according to the running state signal, and combine the belt slip related feature with the theoretical belt speed and the instantaneous unit length material weight to calculate an actual running speed of the conveyor belt according to a pre-labeled mapping relationship.

[0064] A conveying capacity calculation module is configured to calculate a corrected instantaneous material conveying capacity based on the actual running speed and the instantaneous unit length material weight, and represent a conveying capacity deviation by a relative difference between the theoretical belt speed and the actual running speed.

[0065] An energy consumption monitoring module is configured to monitor the real-time power of the driving motor and calculate the additional energy consumption in the monitoring period according to the reference power established under the no-slip condition;

[0066] A cost accounting module is configured to monitor the running state and energy consumption of the downstream associated equipment, the fuel consumption of the downstream coal-consuming equipment, and the environmental protection agent consumption of the flue gas treatment device, and account for the downstream compensation cost caused by the conveying capacity deviation in combination with economic parameters;

[0067] A report generation module is configured to aggregate the additional energy consumption and the downstream compensation cost and calculate the cumulative economic loss, and generate an economic benefit and energy consumption loss report;

[0068] A decision guidance module is configured to output a maintenance suggestion or a dispatch correction instruction when the cumulative economic loss and / or the conveying capacity deviation reaches a preset threshold. Advantages

[0069] The present application effectively solves the problem that the slight slip of the belt caused by the uneven wear of the rubber layer of the driving drum in the prior art cannot be detected in time, resulting in systematic error of the instantaneous conveying capacity calculation system, downstream material shortage and manual intervention, and ultimately affecting the optimization effect of coal resource scheduling and clean and efficient utilization. The present application introduces a calculation mechanism of the actual running speed to make up for the lack of direct measurement of the actual speed of the belt in the existing system, so as to accurately identify and quantify the conveying capacity deviation. In addition, by comprehensively considering the additional energy consumption and the downstream compensation cost, the subtle faults in the physical layer are closely related to the economic benefits, and the quantitative evaluation of the implicit loss is realized. Based on this, the system can generate a precise economic benefit and energy consumption loss report, and output a maintenance suggestion or a dispatch correction instruction in time, thereby avoiding unplanned manual intervention, optimizing the overall scheduling plan, significantly improving the fine management level and economic benefits of coal resource scheduling, and realizing the clean and efficient utilization of coal. BRIEF DESCRIPTION OF DRAWINGS

[0070] Figure 1 A flowchart of a coal resource scheduling method provided by the present application.

[0071] Figure 2 A program block diagram of a coal resource scheduling system provided by the present application.

[0072] In the figure: 1, signal acquisition module; 2, speed calculation module; 3, conveying capacity calculation module; 4, energy consumption monitoring module; 5, cost accounting module; 6, report generation module; 7, decision guidance module. DETAILED DESCRIPTION

[0073] The technical solutions in the present application will be described clearly and completely in the present application combined with the drawings in the present application. Obviously, the described embodiments are only some of the embodiments of the present application, not all the embodiments. The components of the present application described and shown in the drawings can be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present application provided in the drawings is not intended to limit the scope of the claimed present application, but only represents selected embodiments of the present application. Based on the embodiments of the present application, all other embodiments obtained by those skilled in the art without creative labor are within the scope of protection of the present application.

[0074] With reference to Figure 1 The present application proposes a coal resource scheduling method, comprising:

[0075] S1000: Obtain the running state signal of the driving drum area of the conveyor belt, the theoretical speed of the driving motor, and the theoretical belt speed converted therefrom, the instantaneous unit length material weight output by the weighing device, and the economic parameters;

[0076] S2000: Obtain the belt slip related features according to the running state signal, combine the belt slip related features with the theoretical belt speed and the instantaneous unit length material weight, and calculate the actual running speed of the conveyor belt according to the pre-calibrated mapping relationship;

[0077] S3000: Calculate the corrected instantaneous material conveying capacity based on the actual running speed and the instantaneous unit length material weight, and represent the conveying capacity deviation by the relative difference between the theoretical belt speed and the actual running speed;

[0078] S4000: Monitor the real-time power of the driving motor, and calculate the additional energy consumption in the monitoring period according to the reference power established under the non-slip working condition;

[0079] S5000: Monitor the running state and energy consumption of the downstream associated equipment, the fuel consumption of the downstream coal-using equipment, and the environmental protection agent consumption of the flue gas treatment device, and calculate the downstream compensation cost caused by the conveying capacity deviation combined with the economic parameters;

[0080] S6000: Aggregate the additional energy consumption and the downstream compensation cost and calculate the cumulative economic loss, and generate an economic benefit and energy consumption loss report;

[0081] S7000: When the cumulative economic loss and / or the conveying capacity deviation reaches the preset threshold, output a maintenance suggestion or a scheduling correction instruction.

[0082] In order to more easily and clearly understand the technical solutions of the present application, some key terms involved therein are first explained.

[0083] The running state signal refers to various physical quantity signals that can reflect the working condition of the driving drum area of the conveyor belt, such as acoustic signals, vibration signals, temperature signals, etc.

[0084] The theoretical speed refers to the speed that the driving motor should reach under ideal no-load or rated load according to its design parameters or control instructions.

[0085] The theoretical belt speed is the expected running speed of the conveyor belt under no-slip conditions, calculated based on the theoretical speed of the driving drum and the drum diameter.

[0086] The instantaneous unit length material weight refers to the weight of the unit length of material carried on the conveyor belt at a certain moment, usually measured in real time by a weighing device.

[0087] The economic parameter refers to various cost and benefit related data involved in the process of coal resource scheduling, such as electricity price, fuel price, environmental protection agent price, equipment maintenance cost, downtime loss cost, etc.

[0088] The belt slip related feature refers to the physical feature that can represent the relative slip phenomenon between the conveyor belt and the driving drum, such as acoustic signals of specific frequency, vibration mode change, etc.

[0089] The mapping relationship refers to the mathematical or logical correspondence between the belt slip related feature, the theoretical belt speed, the instantaneous unit length material weight and the actual running speed of the conveyor belt, which can be obtained through experimental calibration, machine learning model training, etc.

[0090] The corrected instantaneous material conveying capacity refers to the actual conveying capacity obtained by correcting the material weight measured by the weighing device after considering the actual running speed of the belt.

[0091] The conveying capacity deviation refers to the difference between the actual conveying capacity of the conveyor belt and the theoretical conveying capacity, usually represented by the relative difference between the actual belt speed and the theoretical belt speed.

[0092] The additional energy consumption refers to the part of the energy consumption of the driving motor that exceeds the normal no-slip working condition due to belt slip and other abnormal conditions.

[0093] The downstream compensation cost refers to the additional cost generated to make up for the impact of the conveying capacity deviation on the running of the downstream equipment or the decrease of the efficiency, such as the increase of fuel consumption, the increase of environmental protection agent consumption, the cost of manual intervention, etc.

[0094] The cumulative economic loss refers to the sum of the additional energy consumption and the downstream compensation cost within a certain monitoring period.

[0095] The preset threshold refers to a critical value of cumulative economic loss and / or deviation of conveying capacity used to determine whether maintenance or dispatching modification measures need to be taken.

[0096] The maintenance suggestion refers to measures such as maintenance, replacement of parts, etc. proposed for equipment failure or performance decline.

[0097] The dispatching modification instruction refers to adjustment of an existing coal resource dispatching plan to adapt to changes in actual conveying capacity.

[0098] The coal resource dispatching method provided in the application takes fine perception of the running state of the conveying belt and multi-dimensional economic benefit evaluation as the core, and realizes intelligent optimization and decision-making for the coal conveying process.

[0099] Specifically, in terms of obtaining the running state signal of the driving drum area of the conveying belt, the theoretical speed of the driving motor and the theoretical belt speed converted therefrom, the instantaneous unit length material weight output by the weighing device, and the economic parameters, the running state signal can be obtained by arranging vibration sensors, temperature sensors, acoustic sensors, etc. in the driving drum area; the theoretical speed is obtained from the motor controller or the equipment nameplate, and the theoretical belt speed is calculated in combination with the drum diameter; the instantaneous unit length material weight is measured in real time by using the electronic belt scale installed on the conveying belt; and the economic parameters can be obtained from the enterprise resource planning (ERP), the market price database or manual input.

[0100] Secondly, belt slip related features are extracted from the running state signal, and are input into a pre-labeled mapping relationship together with the theoretical belt speed and the instantaneous unit length material weight to calculate the actual running speed of the conveying belt. When the running state signal is a vibration signal, frequency spectrum analysis can be performed to extract the energy change of a specific frequency band as a belt slip related feature; when the running state signal is an acoustic signal, friction sound or howling sound features related to slip can be identified. The above features, together with the theoretical belt speed and the instantaneous unit length material weight, are input into a machine learning model (such as a support vector machine or a neural network) trained by historical data to obtain real-time estimation of the actual running speed.

[0101] Further, the corrected instantaneous material conveying capacity is calculated based on the actual running speed and the instantaneous unit length material weight, and the relative difference between the theoretical belt speed and the actual running speed is used to represent the conveying capacity deviation. For example, if the instantaneous unit length material weight is X tons / meter and the actual running speed is Y meters / second, the corrected instantaneous material conveying capacity is X × Y tons / second; and the conveying capacity deviation can be calculated as: deviation = (theoretical belt speed actual belt speed) theoretical belt speed × 100%.

[0102] Further, the real-time power of the driving motor is monitored, and the additional energy consumption in the monitoring period is calculated according to the reference power established under the no-slip condition. Specifically, the motor power data is collected under the normal condition of no slip to establish a reference power model that can change with the load and the theoretical speed; and the additional energy consumption is calculated as the positive cumulative part of the difference between the real-time power and the reference power in the monitoring period.

[0103] Further, the running state and energy consumption of the downstream associated equipment, the fuel consumption of the downstream coal-using equipment, and the environmental protection agent consumption of the flue gas treatment device are monitored, and the downstream compensation cost caused by the deviation of the conveying capacity is calculated in combination with the economic parameters. When there is a shortage of material, the downstream equipment may need to extend the running time, increase the load, or change the working condition to make up for the material, resulting in additional consumption of electric energy, fuel, and environmental protection agent; in combination with the electricity price, fuel price, and agent price, the downstream compensation cost can be quantified.

[0104] Subsequently, the additional energy consumption and the downstream compensation cost are summarized and the cumulative economic loss is calculated to form an economic benefit and energy consumption loss report. The report can show the details of the energy consumption loss, the composition of the compensation cost, and the time trend of the cumulative economic loss, facilitating management decision-making.

[0105] Finally, when the cumulative economic loss and / or the deviation of the conveying capacity reaches a preset threshold, a maintenance suggestion or a dispatch correction instruction is output. For example, if the deviation of the conveying capacity continuously exceeds a certain percentage, or the cumulative economic loss reaches a certain amount, a maintenance suggestion can be generated to prompt the inspection or repair of the driving drum of the conveyor belt; if the maintenance level has not been reached but has affected the downstream dispatch, a dispatch correction instruction can be generated, such as adjusting the subsequent material conveying plan or optimizing the running order of the downstream equipment, to reduce the economic loss.

[0106] In another embodiment of the present application, it is further proposed that S2000 comprises:

[0107] S2100: The running state signal comprises a multi-channel acoustic data stream collected by a multi-channel acoustic sensor array arranged in the area of the conveyor belt driving drum, the multi-channel acoustic data stream is spatially located, and acoustic characteristic calibration is performed to obtain the spatial direction of the slip sound source and at least one noise source and the calibration parameters for spatial filtering;

[0108] S2200: Based on the spatial direction and the calibration parameters, the multi-channel acoustic data stream is spatially filtered to enhance the signal from the direction of the slip sound source and attenuate the signal from the direction of the noise source, to obtain a focused signal;

[0109] S2300: Extracting a belt slip-related feature from the focused signal;

[0110] S2400: combine the belt slip related features with the theoretical belt speed and the instantaneous unit length material weight, and calculate the actual running speed of the conveying belt according to a mapping relationship.

[0111] Specifically, the multi-channel acoustic sensor array is formed by arranging a plurality of acoustic sensors in a linear array, a circular array or a planar array, etc. geometry configuration, for simultaneously collecting acoustic signals at different spatial positions in the driving drum area to obtain spatial information of the sound source. The multi-channel acoustic data stream is the multi-channel digital or analog acoustic signals collected by the array in a continuous time period.

[0112] Among them, the spatial positioning of the multi-channel acoustic data stream is to determine the spatial position or direction of the slip sound source and the noise source relative to the array by using the time difference of arrival TDOA, generalized cross-correlation GCC or beamforming array signal processing method. The acoustic characteristic calibration is to analyze and parameterize the frequency spectrum, energy, directivity and other acoustic properties to form reusable acoustic fingerprints or models to obtain the spatial direction of the slip sound source and at least one noise source and the calibration parameters for spatial filtering.

[0113] In practical applications, the spatial filtering processing of the multi-channel acoustic data stream based on the spatial direction and the calibration parameters means that a spatial filter is designed according to the known spatial direction and acoustic characteristics by using beamforming, null steering or independent component analysis ICA technology, selectively enhances the signal from the slip sound source direction and suppresses the interference from the noise source direction, and outputs a focused signal with significantly improved signal-to-noise ratio.

[0114] Further, extracting the belt slip related features from the focused signal is to identify and quantify the indicators directly related to the slip in the time domain, frequency domain or time-frequency domain, such as the energy of specific frequency, harmonic structure, transient impact sound and the spectral characteristics and time-varying mode of friction sound. The belt slip related features are combined with the theoretical belt speed and the instantaneous unit length material weight, and the actual running speed of the conveying belt is calculated according to the pre-calibrated mapping relationship; the mapping relationship is established by experiment, simulation or historical data training, and is used to quantify the nonlinear relationship between the features and the actual speed.

[0115] The scheme of the present application introduces spatial dimension information through the multi-channel acoustic sensor array, making spatial positioning and acoustic characteristic calibration possible; the spatial filtering processing implemented accordingly obtains a focused signal with high signal-to-noise ratio, significantly reducing the interference of environmental noise on feature extraction. The belt slip related features obtained from the focused signal are more pure, and combined with the mapping relationship, the actual running speed can be more accurately calculated, providing reliable input for the corrected instantaneous material conveying capacity, additional energy consumption evaluation and downstream compensation cost accounting, overcoming the limitations of traditional methods in stable acquisition of slip features in complex noise environment.

[0116] In some preferred embodiments, the following are described by a specific example:

[0117] Take a large coal mine conveyor belt system as an example: there are crushers, ventilators and other auxiliary equipment near the driving drum area, and a multi-channel acoustic sensor array composed of eight microphones is laid out to continuously collect multi-channel acoustic data streams. First, spatial positioning and acoustic characteristic calibration are performed, the spatial directions of the crushers, ventilators, auxiliary equipment relative to the array are determined by the time difference of arrival, and the respective frequency spectrum and energy distribution are extracted to obtain calibration parameters, while the typical spatial direction and acoustic characteristics of the belt slip sound source are pre-calibrated. Subsequently, based on the determined spatial direction and calibration parameters, spatial filtering processing is implemented, beamforming is used to align the main lobe to the slip sound source direction, and nulls are set in the directions of the crushers, ventilators, and auxiliary equipment to maximize the enhancement of the slip sound source signal and the suppression of interference. Finally, belt slip-related features such as friction sound specific frequency band energy and periodic impact sound intensity are extracted from the focused signal, combined with theoretical belt speed and instantaneous unit length material weight, and input into the mapping relationship model to accurately calculate the actual running speed; when the friction sound energy rises, the model gives the decline amplitude of the actual belt speed relative to the theoretical belt speed. In the strong noise environment of the crusher and the ventilator, the slip can still be stably identified and the reliable actual running speed can be obtained, providing reliable data support for coal resource scheduling.

[0118] In another embodiment of the present application, further proposed is that the multi-channel acoustic data stream is spatially positioned, and acoustic characteristic calibration is performed to obtain the spatial direction of the slip sound source and at least one noise source and the calibration parameters for spatial filtering, specifically comprising:

[0119] S2110: According to the multi-channel acoustic data stream combined with the running state of the conveyor belt, the start-stop state of the dust suppression system, and the running state information of the noise-related equipment, identify specific time periods when only the slip sound source or only the noise source exists;

[0120] S2120: In the specific time period, the current spatial direction of the slip sound source and at least one noise source relative to the multi-channel acoustic sensor array is determined using the time difference of arrival and / or the generalized cross-correlation algorithm;

[0121] S2130: In the specific time period, the frequency spectrum characteristics, energy distribution and directivity characteristics of the slip sound source and the noise source are analyzed to obtain their current calibration parameters;

[0122] S2140: When a new noise source or a change in the spatial direction of an existing noise source is detected, spatial positioning and acoustic characteristic calibration of the new or changed noise source are triggered to update the spatial direction and calibration parameters;

[0123] S2150: According to the real-time determined spatial direction and the calibration parameters, the parameters of the spatial filtering processing, including the beamforming weights and / or the nulling directions, are updated to maintain the signal-to-noise ratio of the focused signals and continuously enhance the signals from the slipping sound source directions and attenuate the signals from the noise source directions.

[0124] Specifically, identifying the specific time period in which only the slipping sound source or only the noise source exists means that the system determines whether the current acoustic environment is a single sound source scene by analyzing the multi-channel acoustic data stream in combination with the conveyor belt running state, the dust suppression system start-stop state, and the running state information of the noise-related equipment. For example, when the conveyor belt is empty and no other equipment is running, and a specific acoustic signal is detected, it can be identified as a specific time period in which only the slipping sound source exists. Conversely, when the conveyor belt is stopped and a certain noise-related equipment is started, it can be identified as a specific time period in which only the noise source exists. This step provides pure samples for subsequent spatial positioning and acoustic characteristic calibration.

[0125] Among them, the Time Difference of Arrival (TDOA) and / or Generalized Cross-Correlation (GCC) algorithm is used to calculate the spatial direction of the sound source relative to the multi-channel acoustic sensor array according to the time difference of the same sound source received by different sensors, combined with the sound speed and array geometry; if necessary, the Beamforming is used to improve the stability of direction estimation. Thus, the spatial direction of the slipping sound source and the noise source is obtained as the basic positioning information.

[0126] In practical applications, the frequency spectrum characteristics, energy distribution, and directivity characteristics of the positioned sound sources are analyzed to obtain the current calibration parameters. The calibration parameters can include center frequency, bandwidth, instantaneous power spectral density, etc. The slipping friction sound usually has continuous energy rise in the mid-high frequency band and bandwidth expansion characteristics with load; the fan noise often has narrow-band fundamental frequency and multiple frequency peak values; the impact sound of the crusher is usually wide-band pulse energy with periodic changes in directivity. The above analysis results are used for parameterization description of subsequent spatial filtering processing.

[0127] Further, when a new noise source is detected or the spatial direction of an existing noise source changes, the system triggers spatial positioning and acoustic characteristic calibration of the new or changed noise source, and updates the spatial direction and calibration parameters in a timely manner to adapt to the dynamic acoustic environment of the industrial site. For example, when a new fan is put into operation, its spatial direction and calibration parameters need to be added immediately; when a mobile device passes near the sensor array, its spatial direction changes over time, and the system needs to automatically reposition and update the parameters.

[0128] Thus, according to the real-time determined spatial direction and calibration parameters, the parameters of the spatial filtering processing, including the beamforming weight and the null direction, are updated to maintain the signal-to-noise ratio of the focused signal and continuously enhance the signal from the slipping sound source direction and attenuate the signal from the noise source direction. In practice, the beam main lobe can be directed to the slipping sound source direction, while the null direction is set in the direction of the ventilator and the crusher; when the intermittent operation of the crusher causes the radiation directivity to change, the null direction is adjusted synchronously in time sequence to avoid noise leakage into the focused signal.

[0129] In some preferred embodiments,

[0130] The multi-channel acoustic sensor array composed of eight microphones is laid out on site in the vicinity of the driving drum area where the crusher, ventilator and other auxiliary equipment are located. In a certain time period, the conveyor belt is running empty and the crusher, ventilator and auxiliary equipment are all shut down, and the system identifies this specific time period as the time period when only the slipping sound source exists; the spatial direction of the slipping sound source relative to the array is determined by TDOA and GCC, and the calibration parameters of the slipping sound source are obtained through spectral feature, energy distribution and directivity feature analysis. Subsequently, the ventilator is started to form a new noise source, and the system immediately triggers the spatial positioning and acoustic characteristic calibration of the noise source, updates its spatial direction and corresponding calibration parameters; accordingly, the parameters of the spatial filtering processing are dynamically adjusted, the beamforming weight is adjusted to direct the main lobe to the slipping sound source direction, and the null direction is set in the direction of the crusher, ventilator and auxiliary equipment. After the updated spatial filtering processing, the focused signal with high signal-to-noise ratio is obtained, from which the belt slipping related features are accurately extracted, and combined with the theoretical belt speed and instantaneous unit length material weight, the actual running speed of the conveyor belt is calculated according to the mapping relationship, which can still maintain the accuracy of the calculation result even in a strong noise environment.

[0131] In another embodiment of the present application, it is further proposed that, according to the real-time determined spatial direction and calibration parameters, the parameters of the spatial filtering processing, including the beamforming weight and / or the null direction, are updated, specifically including:

[0132] S2151: Real-time monitoring of the moving track and transient acoustic features of the noise source, the moving track being determined by the change of the spatial direction of the noise source relative to the multi-channel acoustic sensor array with time, and the transient acoustic features being determined by the time change of the spectrum and energy distribution of the multi-channel acoustic data stream;

[0133] S2152: Based on the moving track, predicting the spatial direction of the noise source in the next time period;

[0134] S2153: Based on the transient acoustic features, predicting the calibration parameters of the noise source in the next time period;

[0135] S2154: based on the predicted spatial direction and calibration parameters, pre-calculate the parameters for spatial filtering processing in the next time period in the current time period, the parameters for spatial filtering processing include beamforming weights and / or null directions.

[0136] Specifically, real-time monitoring of the moving trajectory and transient acoustic characteristics of the noise source refers to the system continuously tracking the position changes and acoustic property changes of each noise source in the environment. The moving trajectory of the noise source refers to the continuous record of its spatial direction relative to the multi-channel acoustic sensor array over time, which can be determined by continuous time difference of arrival TDOA or generalized cross-correlation GCC at each time to connect the azimuth and elevation angles as a trajectory. The transient acoustic characteristics are determined by the changes of the spectrum and energy distribution of the multi-channel acoustic data stream in the time dimension, which can be extracted by short-time Fourier transform STFT or other time-frequency analysis methods to extract the main frequency, bandwidth and energy properties that change over time.

[0137] Further, based on the moving trajectory, the spatial direction of the noise source in the next time period is predicted. Kalman filtering, particle filtering or machine learning models based on historical trajectory data (such as recurrent neural network RNN) can be used to infer the displacement trend in a short time window according to the existing motion pattern, so as to obtain the prediction result of the spatial direction in the next time period.

[0138] At the same time, based on the transient acoustic characteristics, the calibration parameters of the noise source in the next time period are predicted. The calibration parameters can include spectral characteristics, energy distribution, directivity pattern, etc., which can be learned by using time series analysis models (such as ARIMA), hidden Markov models HMM or deep learning models to learn the historical change law and give the acoustic property prediction at future time.

[0139] Thus, based on the predicted spatial direction and calibration parameters, the parameters for spatial filtering processing in the next time period are pre-calculated in the current time period. The parameters for spatial filtering processing include beamforming weights and / or null directions: the former is used to adjust the amplitude and phase of each channel signal to form a sound beam pointing to the target direction, and the latter is used to form a null in the direction of the noise source to maximize the attenuation of the signal in that direction. The purpose of pre-calculation is to directly apply the optimal parameters when the next time period comes, avoiding the lag caused by real-time detection and recalculation.

[0140] In a typical scenario, a mobile cleaning device is the main noise source in the area of the conveyor drive drum. The device moves along the side of the conveyor belt and its working mode changes periodically, resulting in transient acoustic characteristics fluctuating over time. The system continuously collects a multi-channel acoustic data stream through a multi-channel acoustic sensor array, monitors its moving trajectory in real time, and smoothes and predicts the spatial direction of the history data using Kalman filtering to obtain the spatial direction in the next few seconds. The system analyzes the frequency spectrum and energy distribution of the device, extracts the transient acoustic characteristics, and uses the long short-term memory network (LSTM) to predict the main frequency and energy changes in the next time period. Based on the above prediction results, the system pre-calculates the beamforming weight and null direction in the next time period at the current time period; when the next time period comes, it is directly applied, so that the sound beam continuously points to the direction of the slipping sound source, and a null is formed in the direction of the cleaning device. Even if the device position and acoustic characteristics change constantly, the parameters of the spatial filtering process are still highly matched with its state, the noise is continuously suppressed, the focused signal remains clear, and the accuracy of the actual running speed calculation of the conveyor belt is ensured.

[0141] In another embodiment of the present application, when the cumulative economic loss and / or the deviation of the conveying capacity reaches a preset threshold, a maintenance suggestion or a dispatch correction instruction is output, including:

[0142] S7100: Based on the economic benefit and energy loss report, calculate the growth rate of the cumulative economic loss and the trend index of the conveying capacity deviation;

[0143] S7200: Establish a planned period cost prediction model based on economic parameters, predict the changes of the unit price of electric energy, the unit price of fuel and the unit price of environmental protection agent in the pre-designed planning period, and obtain a cost prediction sequence;

[0144] S7300: Based on the cost prediction sequence and the growth rate and trend index, respectively estimate the total cost of the planning period under a plurality of scenarios, including the predicted additional energy consumption and downstream compensation cost, and the maintenance cost and downtime loss cost;

[0145] S7400: Determine the candidate maintenance time point and the dynamic adjustment amount of the preset threshold that minimizes the difference of the total cost of the planning period, define the maintenance condition, and the maintenance condition at least includes: at any candidate maintenance time point, the difference of the total cost of the planning period of the plurality of scenarios is not less than the preset maintenance trigger threshold, and / or the growth rate of the cumulative economic loss and the trend index of the conveying capacity deviation meet the preset joint threshold;

[0146] S7500: When the maintenance condition is met, generate a maintenance suggestion;

[0147] S7600: When the maintenance condition is not met, a scheduling correction instruction is formed, and the scheduling correction instruction at least includes: adjusting the operation time and start-stop sequence of the downstream associated equipment, adjusting the fuel consumption plan of the downstream coal-consuming equipment, and adjusting the environmental protection agent adding strategy of the flue gas treatment device.

[0148] Specifically, after receiving the economic benefit and energy consumption loss report, the cumulative economic loss is first subjected to time series analysis to calculate its growth rate; methods such as moving average, exponential smoothing or regression analysis can be used to quantify the loss accumulation speed. At the same time, the historical data of the transportation capacity deviation is subjected to trend analysis, and the future trend and amplitude are predicted through linear regression or nonlinear fitting to obtain the trend index of the transportation capacity deviation, which provides input for subsequent prediction and decision-making.

[0149] Further, combined with economic parameters (including historical market prices, policy changes, seasonal factors, etc.), a planned period cost prediction model is established to predict the changes of the unit price of electric energy, the unit price of fuel and the unit price of environmental protection agent in the pre-planned period, and a cost prediction sequence is formed. The model can use time series methods (such as ARIMA, Prophet), machine learning methods (such as support vector regression, neural network) or expert rules.

[0150] On this basis, in order to cope with uncertainties, in a plurality of preset scenarios (such as optimistic, neutral, and pessimistic), the total cost of the planned period is estimated based on the cost prediction sequence, the growth rate of the cumulative economic loss and the trend index of the transportation capacity deviation. The total cost of the planned period takes into account the expected additional energy consumption, downstream compensation cost, as well as maintenance cost and downtime loss cost, to comprehensively evaluate the economic impact of different decision paths.

[0151] As a preferred embodiment, the candidate maintenance time point with the smallest difference is found, and the preset threshold is dynamically adjusted accordingly; at the same time, the maintenance condition is defined, which at least includes: at any candidate maintenance time point, the difference of the total cost of the planned period in multiple scenarios is not less than the preset maintenance trigger threshold, and / or the growth rate of the cumulative economic loss and the trend index of the transportation capacity deviation meet the preset joint threshold. The former guarantees economic rationality, and the latter guarantees technical necessity.

[0152] When the maintenance condition is met, a maintenance suggestion is generated, such as performing maintenance, replacing parts or implementing preventive maintenance on a specific date; when the maintenance condition is not met, a scheduling correction instruction is formed, which at least includes: adjusting the operation time and start-stop sequence of the downstream associated equipment, adjusting the fuel consumption plan of the downstream coal-consuming equipment, and adjusting the environmental protection agent adding strategy of the flue gas treatment device, so as to reduce the total cost under the premise of meeting the environmental protection and capacity constraints.

[0153] By accumulating the growth rate of economic loss and the trend index of the deviation of the conveying capacity (for example, the monthly growth rate is fitted by exponential smoothing / rolling regression; the trend index is predicted by linear / nonlinear fitting or slope), the risk and loss trend are identified in advance; by the planned period cost prediction model and multi-scenario analysis, the future fluctuations in the unit prices of electricity, fuel and environmental protection agents are taken into consideration; the planned period total cost difference is taken as the optimization target and the threshold is dynamically adjusted, so that the decision is more economically stable and can balance the maintenance cost, downtime loss and additional energy consumption caused by continuous operation and downstream compensation cost; when the maintenance condition is not met, the dispatching correction instruction is used to realize flexible optimization on the operation side, thereby minimizing the loss and maintaining continuous production.

[0154] In a typical scenario, a certain conveying system monitors that the deviation of the conveying capacity continues to rise, and the cumulative economic loss continues to increase. The system first calculates that the growth rate of the cumulative economic loss is 5% per month, and the trend index of the deviation of the conveying capacity predicts that it will reach 20% after three months. Then, a planned period cost prediction model is established for a six-month planning period: the unit price of electricity is expected to rise by 10%, the unit price of fuel is basically stable, and the unit price of environmental protection agents is expected to fall by 5%. Under the three scenarios of optimism, neutrality and pessimism, the total cost of the planning period is estimated to be 1 million yuan, 1.2 million yuan and 1.5 million yuan respectively. Taking the scenario total cost difference as the optimization target, the algorithm determines the third month as the candidate maintenance time point with the smallest difference, and dynamically adjusts the preset threshold accordingly; at the same time, the maintenance condition is set: if the maintenance is carried out at the end of the third month, the scenario total cost difference is not less than the preset maintenance trigger threshold (such as 200,000 yuan), and the growth rate of the cumulative economic loss and the trend index of the deviation of the conveying capacity meet the preset joint threshold. If the maintenance condition is met, the system outputs the maintenance suggestion, such as carrying out a comprehensive overhaul of the driving roller of the conveying belt and replacing the worn parts at the end of the third month; if the maintenance condition is not met, a dispatching correction instruction is generated, such as adjusting the operation time and start-stop sequence of the downstream associated equipment, optimizing the fuel consumption plan of the downstream coal-using equipment, and optimizing the environmental protection agent dosing strategy of the flue gas treatment device under the condition of meeting the standard. Through this process, even if maintenance is not carried out immediately, the economic loss can be reduced and the production can be stabilized through fine scheduling.

[0155] In another embodiment of the present application, it is further proposed to monitor the moving track and transient acoustic characteristics of the noise source in real time, including:

[0156] S2151-1: In the initialization stage or when the noise source changes are detected, spatial positioning and acoustic characteristic calibration are performed for each noise source respectively to obtain the initial spatial direction and calibration parameters thereof;

[0157] S2151-2: A multi-target sound source separation algorithm is applied, the initial spatial direction and calibration parameters of each noise source are combined, and the signal components corresponding to each noise source in the multi-channel acoustic data stream are separated into independent acoustic signals corresponding to each noise source.

[0158] S2151-3: Time-frequency analysis is performed on each separated independent acoustic signal to extract its transient acoustic features.

[0159] S2151-4: Using each independent acoustic signal, the current spatial direction of each noise source relative to the multi-channel acoustic sensor array is determined and continuously updated using a time difference of arrival (TDOA) and / or generalized cross-correlation (GCC) algorithm to construct the moving trajectory of each noise source.

[0160] Specifically, the initialization stage refers to the first identification and parameter acquisition of existing noise sources in the environment when the system is first started or undergoes large-scale configuration updates. Noise source changes refer to the discovery of new noise sources or significant changes in the acoustic characteristics and location of existing noise sources through acoustic scene analysis or anomaly detection. Spatial positioning and acoustic characteristic calibration are performed for each noise source to obtain its spatial direction relative to the multi-channel acoustic sensor array and calibration parameters such as frequency spectrum and energy distribution, providing accurate initial values for subsequent sound source separation and tracking.

[0161] The multi-target sound source separation algorithm is used to separate the signal components corresponding to each noise source in the multi-channel acoustic data stream into independent acoustic signals. Combining the initial spatial direction and calibration parameters of each noise source, independent component analysis (ICA), non-negative matrix factorization (NMF), or beamforming-based separation techniques can be selected to improve the effectiveness and stability of separation.

[0162] Time-frequency analysis is performed on each separated independent acoustic signal to reveal its energy distribution and variation in the time and frequency dimensions, and to extract transient acoustic features. The transient acoustic features can include short-time energy, zero-crossing rate, spectral centroid, spectral bandwidth, mel-frequency cepstral coefficient (MFCC), etc., which are used to characterize the transient properties of noise sources and support subsequent identification and prediction.

[0163] Using each independent acoustic signal, the current spatial direction of each noise source relative to the multi-channel acoustic sensor array is accurately determined using a time difference of arrival (TDOA) and / or generalized cross-correlation (GCC) algorithm. TDOA calculates the sound source position by measuring the time difference between multiple sensors, and GCC estimates the time delay through cross-correlation peaks and indirectly obtains TDOA. Continuous updating of the spatial direction can construct a moving trajectory that evolves over time, achieving dynamic tracking of noise sources.

[0164] The scheme of the present application integrates the initialization / change detection, spatial positioning and acoustic characteristic calibration, multi-target sound source separation, time-frequency analysis to extract transient acoustic characteristics, TDOA / GCC continuous positioning and moving track construction into an integrated process: on the one hand, it ensures cleaner independent acoustic signal separation, and on the other hand, it enables the spatial direction and moving track of the noise source to be updated in real time and accurately. Thus, the subsequent spatial filtering process can obtain timely and accurate input parameters, thereby improving the effect of slip sound source enhancement and noise attenuation, and further improving the accuracy and robustness of the actual running speed calculation of the conveyor belt, and optimizing the overall efficiency and reliability of the coal resource scheduling.

[0165] In another embodiment of the present application, it is further proposed that S2151-2 comprises:

[0166] S2151-21: in combination with the multi-channel acoustic data stream, the running state of the conveyor belt, the corrected instantaneous material conveying capacity and the real-time power of the driving motor, it is determined whether the acoustic scene has a transient change;

[0167] S2151-22: when it is determined that the acoustic scene has a transient change, sound source activity detection is performed on the multi-channel acoustic data stream to identify a potential active sound source region;

[0168] S2151-23: the acoustic characteristics of the potential active sound source region are compared with the preset slip sound source fingerprint library and the environmental background noise fingerprint library to obtain a category determination and / or a similarity index;

[0169] S2151-24: for a sound source region with a similarity to the slip sound source fingerprint library or the environmental background noise fingerprint library greater than a preset similarity threshold, a lower noise source confidence weight is given to it in the multi-target sound source separation algorithm according to a preset weight rule and / or it is excluded from the noise source separation target;

[0170] S2151-25: for a sound source region with a similarity to the slip sound source fingerprint library or the environmental background noise fingerprint library less than a preset similarity threshold, and a similarity to at least one noise source fingerprint in the preset noise source fingerprint library greater than a preset noise threshold, it is added to the noise source separation target as a new noise source, and a higher noise source confidence weight is given to it in the multi-target sound source separation algorithm according to a preset weight rule;

[0171] S2151-26: during the execution of the multi-target sound source separation algorithm, the separation strategy is dynamically adjusted according to the noise source confidence weight;

[0172] S2151-27: the separated independent acoustic signals are cross-verified, which includes comparison with historical noise source moving tracks, acoustic fingerprints and / or running state information of noise-related equipment to confirm whether it is a real noise source.

[0173] Specifically, determining whether the acoustic scene has a transient change refers to comprehensively analyzing multi-dimensional data such as the multi-channel acoustic data stream, the running state of the conveying belt, the corrected instantaneous material conveying amount, and the real-time power of the driving motor, to identify whether a significant and unexpected change occurs in the current acoustic environment, such as the appearance of a new noise source, a sudden mutation or disappearance of the intensity of an existing noise source, and the like. The purpose is to timely trigger a more refined sound source processing flow.

[0174] Sound source activity detection refers to locating a specific time-frequency region or a spatial region where sound source activity exists from the multi-channel acoustic data stream by using energy detection, spectrum analysis or machine learning methods, so as to determine the potential active sound source region. The purpose is to narrow the analysis range and focus on the signal part that may contain a noise source or a slipping sound source.

[0175] The slipping sound source fingerprint library, the environmental background noise fingerprint library and the noise source fingerprint library are pre-established databases that collect acoustic feature sets of various typical sound sources, such as frequency spectrum, energy distribution and time domain waveform characteristics. Comparing the acoustic characteristics of the potential active sound source region with the above fingerprint libraries can obtain a category determination and / or a similarity index, so as to preliminarily determine the sound source type and provide a basis for subsequent allocation of noise source confidence weight.

[0176] The noise source confidence weight is used to guide the multi-target sound source separation algorithm. A higher weight indicates that the region is more likely to be a real noise source, and the separation algorithm will tend to separate it as an independent noise source; a lower weight or being excluded from the separation target indicates that the region is more likely to correspond to a slipping sound source or unimportant background noise. The purpose is to optimize the separation performance and avoid mistakenly including a slipping sound source or background noise in complex separation.

[0177] Dynamic adjustment of separation strategy refers to adjusting internal parameters and processing flow in the multi-target sound source separation algorithm during execution, such as adjusting the beamforming direction, filter coefficients or objective function in the iterative optimization process, according to the real-time updated noise source confidence weight, to adapt to the changing acoustic scene and improve the separation accuracy and robustness.

[0178] Cross-validation refers to carrying out multi-dimensional and multi-source verification on each independent acoustic signal obtained by separation, including comparing with historical noise source moving track, acoustic fingerprint and / or running state information of noise-related equipment, to further confirm whether it corresponds to a real noise source, thereby reducing the risk of misjudgment and omission.

[0179] In some preferred embodiments:

[0180] If new high-frequency noise suddenly appears in the area of the drive drum of the conveyor belt, the system will combine multi-channel acoustic data streams, the running state of the conveyor belt such as empty or full, the corrected instantaneous material conveying capacity, and the real-time power of the drive motor to determine whether it is a transient acoustic event; once confirmed, immediately perform sound source activity detection to locate the potential active sound source area of the high-frequency noise. Then, compare its acoustic characteristics with the slip source fingerprint library, the environmental background noise fingerprint library, and the various noise source fingerprint library: if the similarity with the slip source fingerprint library or the environmental background noise fingerprint library is relatively high, for example, higher than the preset similarity threshold, then give a lower noise source confidence weight or exclude it; if the similarity with the above two types of fingerprints is relatively low and the similarity with a certain specific noise source fingerprint, for example, the noise fingerprint of a certain fan or vibrating screen, is relatively high, for example, higher than the preset noise threshold, then it will be added to the separation target as a new noise source and given a higher weight. The multi-target sound source separation algorithm dynamically adjusts the beamforming direction and intensity accordingly, more effectively separating the high-frequency noise; after separation, cross-verification is performed with the historical noise source moving track, acoustic fingerprint, and the start-stop state and power consumption of the corresponding equipment to finally confirm whether the high-frequency noise is the real noise source of the corresponding fan or vibrating screen. Through this path, the accuracy of noise source separation is ensured, and the slip source is avoided from being misjudged as a noise source, thereby improving the reliability of the belt slip detection.

[0181] In another embodiment of the present application, the step of cross-verifying the separated independent acoustic signals further comprises:

[0182] S2151-271: Obtain the running state information of the noise-related equipment, the running state information including the start-stop state and power consumption of the noise-related equipment;

[0183] S2151-272: According to the preset multi-dimensional verification rule set, cross-verify the transient acoustic characteristics, the moving track, and the running state information of the noise-related equipment to obtain a cross-verification judgment result;

[0184] S2151-273: According to the cross-verification judgment result, confirm whether each independent acoustic signal corresponds to a real noise source, and correct the misjudgment or missed judgment for the situation that does not meet the judgment condition.

[0185] Specifically, the running state information of the noise-related equipment refers to the real-time working data of the equipment related to the potential noise source, and the running state information can include the start-stop state and power consumption of the noise-related equipment. By obtaining these information, external evidence can be provided for the authenticity judgment of the acoustic signal. For example, a certain acoustic signal is identified as the noise of a certain fan, but the fan is in shutdown or abnormal power consumption, which needs to be further verified.

[0186] The multi-dimensional verification rule set is a set of preset logical judgment rules for comprehensive analysis of the transient acoustic characteristics, movement trajectory, and running state information of the noise-related equipment. The rules can be formed based on expert experience, historical data analysis, or machine learning models, with the goal of identifying the correlations and inconsistencies between multi-dimensional information. For example, the rules can stipulate that when the acoustic fingerprint matches a known noise source and the corresponding equipment power consumption is within the normal range, it is determined to be a real noise source; when the acoustic fingerprint matches but the equipment is in a shutdown state, it is determined to be an abnormal situation and enters subsequent processing.

[0187] In actual applications, according to the cross-validation determination result, it is confirmed whether each independent acoustic signal corresponds to a real noise source, and the misjudgment or missed judgment of the situation that does not meet the determination condition is corrected. That is, when there is ambiguity, contradiction or error in the determination, the system does not simply accept or reject, but starts a correction mechanism to improve the recognition accuracy and robustness.

[0188] The scheme of the present application solves the problem of inaccurate judgment of traditional cross-validation in complex industrial environments by introducing a multi-dimensional verification rule set and a misjudgment or missed judgment correction mechanism. First, the running state information of the noise-related equipment, such as start-stop state and power consumption, is obtained to provide key non-acoustic dimensional data. Second, the multi-dimensional verification rule set is used to make comprehensive logical judgments on the transient acoustic characteristics, movement trajectory, and equipment running state, avoiding the bias caused by a single dimension. For example, when the acoustic signal matches the fingerprint of a known noise source but the corresponding equipment is not running or has abnormal power, the rules can identify the inconsistency and trigger the correction process. Thus, based on the cross-validation determination result, the system can not only confirm the real noise source, but also correct the situation that does not meet the determination condition, such as further analysis or entering the abnormal processing flow, significantly improving the accuracy and reliability of noise source identification and avoiding scheduling decision errors caused by misjudgment or missed judgment.

[0189] In some preferred embodiments, the multi-channel acoustic sensor array of the conveyor drive drum area detects a new independent acoustic signal. Preliminary analysis shows that the transient acoustic characteristics of the signal are highly matched with a preset noise fingerprint library of a certain fan, but the noise-related equipment operation state information shows that the fan is currently in a shutdown state. If only acoustic fingerprint matching is used, it is easy to make a false judgment. The scheme of the present application will cross-verify the logical judgment according to the multi-dimensional verification rule set: when the fingerprint matches and the equipment is shutdown, the independent acoustic signal is marked as a to-be-confirmed state; then abnormal mode analysis is performed to compare the correlation between the transient acoustic characteristics, the moving track and the historical operation state information of the fan within a preset time window. For example, if the historical data shows that the fan has never produced such an acoustic signal when it is shutdown, the independent acoustic signal can be marked as a non-real noise source; otherwise, if the historical data shows that the fan occasionally produces similar noise due to inertia or failure when it is in a shutdown state, and the current acoustic characteristics match the historical abnormal mode, it can be confirmed as a real noise source and trigger the corresponding maintenance suggestion. When the fan resumes normal operation, the system will re-perform cross-verification logical judgment on the independent acoustic signal in the to-be-confirmed state according to the multi-dimensional verification rule set, and update the confirmation conclusion of whether it is a real noise source, to ensure the dynamic accuracy of the identification result.

[0190] In another embodiment of the present application, it is further proposed that, according to the cross-verification judgment result, it is confirmed whether each independent acoustic signal corresponds to a real noise source, and the scheme for correcting false judgments or missed judgments in the case of not meeting the judgment conditions, specifically includes:

[0191] A1: When the cross-verification judgment result indicates that the transient acoustic characteristics of the independent acoustic signal match the preset noise source fingerprint library, but the operation state information of the noise-related equipment is inconsistent or abnormal, the corresponding independent acoustic signal is marked as a to-be-confirmed state;

[0192] A2: For the to-be-confirmed state, abnormal mode analysis is performed to compare the historical correlation between the transient acoustic characteristics, the moving track and the operation state information of the noise-related equipment within a preset time window, to obtain a historical correlation analysis result;

[0193] A3: When the historical correlation analysis result shows that the independent acoustic signal appears when the noise-related equipment is abnormal or not started, and matches the transient acoustic characteristics collected at the end time of the preset time window, the independent acoustic signal is confirmed as a real noise source;

[0194] A4: When the historical correlation analysis result shows that the independent acoustic signal does not appear when the noise-related equipment is abnormal or not started, and / or its transient acoustic characteristics do not match the transient acoustic characteristics collected at the end time of the preset time window, the independent acoustic signal is marked as a non-real noise source;

[0195] A5: When the running state information of the noise-related device returns to normal, re-cross-verify the independent acoustic signal in the pending confirmation state according to the multi-dimensional verification rule set, and update the confirmation conclusion of whether it is a real noise source.

[0196] Specifically, when the cross-verification result indicates that the transient acoustic characteristics of the independent acoustic signal match the preset noise source fingerprint library, but the running state information of the noise-related device is inconsistent or abnormal (for example, the device is not started but its characteristic acoustic signal is detected, or the device is running but the acoustic characteristics do not match), the system does not immediately draw a conclusion, but marks the independent acoustic signal as a pending confirmation state, to avoid misjudgment caused by transient abnormalities or information mismatch.

[0197] For the pending confirmation state, the system performs abnormal mode analysis. Within a preset time window, the historical correlation between the transient acoustic characteristics, the movement trajectory, and the running state information of the noise-related device is compared and evaluated. For example, whether the signal has a stable corresponding relationship with the start-stop state of the corresponding device in the past period, or whether it appears frequently during the abnormal stage of the device, so as to more comprehensively judge the signal authenticity.

[0198] When the historical correlation analysis result shows that the independent acoustic signal appears when the noise-related device is abnormal or not started, and matches the transient acoustic characteristics collected at the end of the preset time window, it is confirmed as a real noise source; otherwise, if the signal does not appear when the device is abnormal or not started, and / or its transient acoustic characteristics do not match the transient acoustic characteristics collected at the end of the preset time window, it is marked as a non-real noise source, to exclude incidental interference or misidentification.

[0199] In addition, when the running state information of the noise-related device returns to normal, the system re-performs cross-verification logical judgment on the independent acoustic signal in the pending confirmation state according to the multi-dimensional verification rule set, updates the confirmation conclusion of whether it is a real noise source, and realizes dynamic review based on state recovery.

[0200] By introducing the pending confirmation state and abnormal mode analysis, and combining historical correlation for judgment, this scheme can effectively reduce the misjudgment and omission risk of traditional cross-verification in the case of inconsistent or abnormal running state information of the noise-related device. Multi-dimensional and time series joint analysis makes the system more accurate in distinguishing real noise sources and interference signals, thereby improving the reliability of belt slip related feature extraction and actual running speed calculation of the conveying belt, providing a more stable data basis for coal resource scheduling, reducing additional energy consumption and downstream compensation cost, and improving overall economic efficiency.

[0201] In some preferred embodiments, the following scenario can be considered. The belt driving drum area monitors a new independent acoustic signal, the transient acoustic characteristics of which are highly matched with the noise fingerprint of a certain fan, but the operating state information shows that the fan is in shutdown or the power consumption is significantly lower. At this time, the system marks this signal as a to-be-confirmed state, and performs abnormal mode analysis within a preset time window: comparing the transient acoustic characteristics of the signal with the moving track, and the historical operating state of the fan. If the historical record shows that such acoustic signals have never occurred during the shutdown stage of the fan, or the current characteristics do not match the transient acoustic characteristics collected at the end of the preset time window, the signal is marked as a non-real noise source, avoiding misjudgment of external interference or sensor transient failure as equipment noise. On the contrary, if the historical record shows that the fan occasionally produces similar noise during abnormal or non-starting, and the current characteristics match the transient acoustic characteristics collected at the end of the preset time window, it is confirmed as a real noise source, and the corresponding maintenance suggestion can be triggered. Thereafter, when the fan resumes normal operation, the system again performs cross-validation logic judgment according to the multi-dimensional verification rule set, and if the acoustic characteristics are consistent with the normal operating state, the confirmation conclusion is updated to ensure the dynamic accuracy of the identification result.

[0202] Reference Figure 2 The present application proposes a coal resource scheduling system, comprising:

[0203] A signal acquisition module 1 is configured to acquire the operating state signal of the belt driving drum area, the theoretical speed of the driving motor and the theoretical belt speed converted therefrom, the instantaneous unit length material weight output by the weighing device, and economic parameters;

[0204] A speed calculation module 2 is configured to obtain belt slip related characteristics according to the operating state signal, combine the belt slip related characteristics with the theoretical belt speed and the instantaneous unit length material weight, and calculate the actual running speed of the belt according to a pre-labeled mapping relationship;

[0205] A conveying capacity calculation module 3 is configured to calculate the corrected instantaneous material conveying capacity based on the actual running speed and the instantaneous unit length material weight, and represent the conveying capacity deviation by the relative difference between the theoretical belt speed and the actual running speed;

[0206] An energy consumption monitoring module 4 is configured to monitor the real-time power of the driving motor, and calculate the additional energy consumption in the monitoring period according to the reference power established under the no-slip condition;

[0207] A cost accounting module 5 is configured to monitor the operating state and energy consumption of the downstream associated equipment, the fuel consumption of the downstream coal-using equipment, and the environmental protection agent consumption of the flue gas treatment device, and calculate the downstream compensation cost caused by the conveying capacity deviation in combination with the economic parameters;

[0208] The report generation module 6 is used for aggregating the additional energy consumption and the downstream compensation cost, and calculating the cumulative economic loss, and generating an economic benefit and energy consumption loss report;

[0209] The decision guidance module 7 is used for outputting a maintenance suggestion or a scheduling correction instruction when the cumulative economic loss and / or the conveying capacity deviation reaches a preset threshold.

[0210] The coal resource scheduling system of the present application realizes fine perception of the conveying belt running state, accurate calculation of the actual conveying capacity, multi-dimensional quantitative evaluation of the economic loss, and intelligent decision support by integrating multiple functional modules.

[0211] The above is only an embodiment of the present application and is not used to limit the protection scope of the present application. For those skilled in the art, the present application can have various changes and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application shall be included in the protection scope of the present application.

Claims

1. A coal resource scheduling method, characterized in that, include: The system acquires the operating status signal of the conveyor belt drive roller area, the theoretical speed of the drive motor, and the theoretical belt speed, the instantaneous weight of material per unit length output by the weighing device, and economic parameters. Based on the operating status signal, belt slippage-related characteristics are obtained. These characteristics are then combined with the theoretical belt speed and the instantaneous weight of material per unit length. The actual operating speed of the conveyor belt is calculated based on a pre-calibrated mapping relationship. The corrected instantaneous material conveying capacity is calculated based on the actual operating speed and the instantaneous weight of material per unit length, and the relative difference between the theoretical belt speed and the actual operating speed is used to characterize the conveying capacity deviation. Monitor the real-time power of the drive motor and calculate the additional energy consumption during the monitoring period based on the baseline power established under no slippage conditions; Monitor the operating status and energy consumption of downstream related equipment, the fuel consumption of downstream coal-using equipment, and the environmental protection agent consumption of flue gas treatment devices, and calculate the downstream compensation cost caused by the deviation in the conveying capacity based on the economic parameters. Summarize the additional energy consumption and the downstream compensation costs, calculate the cumulative economic loss, and generate an economic benefit and energy loss report; When the cumulative economic loss and / or the deviation in the transport capacity reaches a preset threshold, a maintenance suggestion or a scheduling correction instruction is output. When the cumulative economic loss and / or the transport capacity deviation reaches a preset threshold, a maintenance suggestion or scheduling correction instruction is output, including: Based on the aforementioned economic benefits and energy loss report, calculate the growth rate of the cumulative economic loss and the trend index of the transmission capacity deviation; A cost forecasting model for the planning period is established based on the aforementioned economic parameters to predict the changes in the unit price of electricity, the unit price of fuel, and the unit price of environmental protection agents during the preset planning period, thereby obtaining a cost forecasting sequence. Under various preset scenarios, the total cost for the planning period is estimated based on the cost prediction sequence, the growth rate, and the trend indicators. The total cost for the planning period includes the expected additional energy consumption, the downstream compensation cost, maintenance costs, and downtime loss costs. Determine the candidate maintenance time point that minimizes the difference in the total cost of the planning period and the dynamic adjustment amount of the preset threshold, define maintenance conditions, the maintenance conditions include at least: at any candidate maintenance time point, the difference in the total cost of the planning period of the multiple scenarios is not less than the preset maintenance trigger threshold, and / or the growth rate of the cumulative economic loss and the trend index of the deviation of the transport capacity meet the preset joint threshold. When the maintenance conditions are met, the maintenance suggestion is generated; When the maintenance conditions are not met, the scheduling correction instruction is generated. The scheduling correction instruction includes at least: adjusting the operating time and start-up / shutdown sequence of the downstream related equipment, adjusting the fuel consumption plan of the downstream coal-using equipment, and adjusting the environmental agent dosing strategy of the flue gas treatment device.

2. The coal resource scheduling method according to claim 1, characterized in that, Based on the operating status signal, belt slippage-related characteristics are obtained. These characteristics are then combined with the theoretical belt speed and the instantaneous weight of material per unit length. The actual operating speed of the conveyor belt is calculated based on a pre-calibrated mapping relationship, including: The operating status signal includes a multi-channel acoustic data stream collected by a multi-channel acoustic sensor array deployed in the conveyor belt drive roller area. The multi-channel acoustic data stream is spatially located and its acoustic characteristics are calibrated to obtain the spatial direction of the slippage sound source and at least one noise source, as well as calibration parameters for spatial filtering. Based on spatial orientation and calibration parameters, the multi-channel acoustic data stream is spatially filtered to enhance the signal from the slippage sound source direction and attenuate the signal from the noise source direction, thereby obtaining a focused signal. Extract the belt slippage-related features from the focused signal; By combining the belt slippage characteristics with the theoretical belt speed and the instantaneous weight of material per unit length, the actual operating speed of the conveyor belt is calculated based on the mapping relationship.

3. The coal resource scheduling method according to claim 2, characterized in that, Spatial localization of the multi-channel acoustic data stream and acoustic characteristic calibration to obtain the spatial orientation of the slippage sound source and at least one noise source, as well as calibration parameters for spatial filtering, including: Based on the multi-channel acoustic data stream combined with the conveyor belt operating status, dust suppression system start / stop status, and noise-related equipment operating status information, specific time periods where only slippage sound sources or only noise sources exist are identified. Within the specified time period, the current spatial orientation of the slippage sound source and at least one noise source relative to the multi-channel acoustic sensor array is determined using the time difference of arrival and / or a generalized cross-correlation algorithm. Within the specified time period, the spectral characteristics, energy distribution, and directional characteristics of the slippage sound source and the noise source are analyzed to obtain their current calibration parameters; When a new noise source is detected or the spatial orientation of an existing noise source changes, spatial localization and acoustic characteristic calibration of the new or changed noise source are triggered to update the spatial orientation and calibration parameters. Based on the real-time determined spatial orientation and calibration parameters, update the parameters of the spatial filtering process, including beamforming weights and / or null orientation, to maintain the signal-to-noise ratio of the focused signal and continuously enhance the signal from the slippage sound source direction and attenuate the signal from the noise source direction.

4. The coal resource scheduling method according to claim 3, characterized in that, Based on the real-time determined spatial orientation and calibration parameters, the parameters of the spatial filtering process are updated, including beamforming weights and / or null directions, including: The movement trajectory and transient acoustic characteristics of the noise source are monitored in real time. The movement trajectory is determined by the change of the spatial orientation of the noise source relative to the multi-channel acoustic sensor array over time, and the transient acoustic characteristics are determined by the temporal changes of the spectrum and energy distribution of the multi-channel acoustic data stream. Based on the movement trajectory, predict the spatial direction of the noise source in the next time period; Based on the transient acoustic characteristics, the calibration parameters of the noise source in the next time period are predicted; Based on the predicted spatial orientation and calibration parameters, the parameters for the next time period are pre-calculated within the current time period. The parameters for the spatial filtering process include the beamforming weights and / or the null orientation.

5. The coal resource scheduling method according to claim 4, characterized in that, Real-time monitoring of the movement trajectory and transient acoustic characteristics of the noise source, including: During the initialization phase or when a change in a noise source is detected, spatial localization and acoustic characteristic calibration are performed for each noise source to obtain its initial spatial orientation and calibration parameters. By applying a multi-target sound source separation algorithm and combining the initial spatial orientation and calibration parameters of each noise source, the signal components corresponding to each noise source in the multi-channel acoustic data stream are separated into independent acoustic signals corresponding to each noise source. Time-frequency analysis was performed on each separated independent acoustic signal to extract its transient acoustic features; Using the independent acoustic signals, and employing time difference of arrival and / or generalized cross-correlation algorithms, the current spatial orientation of each noise source relative to the multi-channel acoustic sensor array is determined and continuously updated to construct the movement trajectory of each noise source.

6. The coal resource scheduling method according to claim 5, characterized in that, A multi-target sound source separation algorithm is applied, combining the initial spatial orientation and calibration parameters of each noise source, to separate the signal components corresponding to each noise source in the multi-channel acoustic data stream into independent acoustic signals corresponding to each noise source, including: By combining the multi-channel acoustic data stream, the conveyor belt operating status, the corrected instantaneous material conveying volume, and the real-time power of the drive motor, it is determined whether a transient change has occurred in the acoustic scene. When it is determined that a transient change has occurred in the acoustic scene, sound source activity detection is performed on the multi-channel acoustic data stream to identify potential active sound source regions; The acoustic characteristics of the potential active sound source region are compared with the preset slip sound source fingerprint database and environmental background noise fingerprint database to obtain category determination and / or similarity index. For a sound source region whose similarity to the slipping sound source fingerprint database or the environmental background noise fingerprint database is greater than a preset similarity threshold, a lower noise source confidence weight is assigned and / or it is excluded from the noise source separation target in the multi-target sound source separation algorithm according to a preset weighting rule. For a sound source region whose similarity to the slipping sound source fingerprint database or the environmental background noise fingerprint database is less than a preset similarity threshold, and whose similarity to at least one noise source fingerprint in the preset noise source fingerprint database is greater than a preset noise threshold, it is added as a new noise source to the noise source separation target, and a higher noise source confidence weight is assigned according to the preset weighting rules in the multi-target sound source separation algorithm. During the execution of the multi-target sound source separation algorithm, the separation strategy is dynamically adjusted according to the confidence weight of the noise source; The separated independent acoustic signals are cross-validated, which includes comparing them with historical noise source movement trajectories, acoustic fingerprints, and / or the operating status information of the noise-associated devices to confirm whether they are real noise sources.

7. The coal resource scheduling method according to claim 6, characterized in that, Cross-validation is performed on each of the separated independent acoustic signals. This cross-validation includes comparing the signals with historical noise source movement trajectories, acoustic fingerprints, and / or the operational status information of the noise-associated devices to confirm whether they are genuine noise sources. Obtain the operating status information of the noise-associated device, the operating status information including the start / stop status and power consumption of the noise-associated device; Based on a preset multi-dimensional verification rule set, cross-verification logic is performed on the transient acoustic features, the movement trajectory, and the operating status information of the noise-related device to obtain the cross-verification judgment result. Based on the cross-validation results, it is confirmed whether each independent acoustic signal corresponds to a real noise source, and misjudgments or omissions are corrected for cases that do not meet the judgment conditions.

8. The coal resource scheduling method according to claim 7, characterized in that, Based on the cross-validation results, it is confirmed whether each independent acoustic signal corresponds to a real noise source, and misjudgments or omissions are corrected for cases that do not meet the judgment criteria, including: When the cross-validation result indicates that the transient acoustic characteristics of the independent acoustic signal match the preset noise source fingerprint database, but the operating status information of the noise-associated device is inconsistent or abnormal, the corresponding independent acoustic signal is marked as pending confirmation. For the unconfirmed state, anomaly pattern analysis is performed. Within a preset time window, the historical correlation between the transient acoustic features, the movement trajectory, and the operating status information of the noise-related device is compared to obtain the historical correlation analysis results. When the historical correlation analysis results show that the independent acoustic signal appears when the noise correlation device is abnormal or not started, and matches the transient acoustic characteristics collected at the end of the preset time window, the independent acoustic signal is confirmed as a real noise source. When the historical correlation analysis results show that the independent acoustic signal did not appear when the noise correlation device was abnormal or not started, and / or its transient acoustic characteristics did not match the transient acoustic characteristics collected at the end of the preset time window, the independent acoustic signal was marked as a non-real noise source. When the operating status information of the noise-associated device returns to normal, the independent acoustic signal in the pending confirmation state is cross-validated again according to the multi-dimensional verification rule set, and the confirmation conclusion of whether it is a real noise source is updated.

9. A coal resource dispatching system, characterized in that, In conjunction with the coal resource scheduling method described in claim 1, the system includes: The signal acquisition module is used to acquire the operating status signal of the conveyor belt drive roller area, the theoretical speed of the drive motor and the theoretical belt speed calculated accordingly, the instantaneous weight of material per unit length output by the weighing device, and economic parameters. The speed calculation module is used to obtain belt slippage related characteristics based on the operating status signal, combine the belt slippage related characteristics with the theoretical belt speed and the instantaneous material weight per unit length, and calculate the actual operating speed of the conveyor belt according to the pre-calibrated mapping relationship. The conveying capacity calculation module is used to calculate the corrected instantaneous material conveying capacity based on the actual operating speed and the instantaneous material weight per unit length, and to characterize the conveying capacity deviation by the relative difference between the theoretical belt speed and the actual operating speed. The energy consumption monitoring module is used to monitor the real-time power of the drive motor and calculate the additional energy consumption during the monitoring period based on the baseline power established under no slippage conditions. The cost accounting module is used to monitor the operating status and energy consumption of downstream related equipment, the fuel consumption of downstream coal-using equipment, and the environmental protection agent consumption of flue gas treatment devices, and calculate the downstream compensation cost caused by the deviation in the conveying capacity in combination with the economic parameters. The report generation module is used to summarize the additional energy consumption and the downstream compensation costs and calculate the cumulative economic loss to generate an economic benefit and energy loss report. The decision guidance module is used to output maintenance suggestions or scheduling correction instructions when the cumulative economic loss and / or the transmission capacity deviation reaches a preset threshold.

Images