System for monitoring operation state of decentralized sewage treatment equipment discharged into river channel

Abstract

The application relates to the technical field of sewage treatment, and discloses a running state monitoring system of a decentralized sewage treatment equipment discharged into a river channel, the system comprising: a data acquisition module for acquiring original acoustic signals and multi-point temperature data in a biological reaction unit of the sewage treatment equipment; a passive monitoring module for extracting acoustic feature vectors and temperature feature vectors from the original acoustic signals and the multi-point temperature data; an active detection control module for sending a controlled disturbance instruction to an actuator of the sewage treatment equipment; a dynamic response analysis module for acquiring the original acoustic signals and the multi-point temperature data; and a cooperative diagnosis module for fusing static abnormality scores and dynamic response features. The application realizes low-cost and high-precision differential diagnosis of different fault types such as biological activity reduction and physical blockage by fusing passive monitoring of acoustic and temperature data and combining dynamic response analysis under active disturbance.

Description

Technical Field

[0001] This invention relates to the field of wastewater treatment technology, specifically to a system for monitoring the operational status of decentralized wastewater treatment equipment discharged into rivers. Background Technology

[0002] Decentralized wastewater treatment equipment, as an important supplement to centralized wastewater treatment plants, is widely used in rural towns, highway service areas, and other residential settings. It plays a crucial role in preventing untreated domestic sewage from being directly discharged into rivers and protecting the aquatic environment. However, the large number of these devices, their dispersed geographical locations, and the fact that they are mostly unattended operations result in extremely high on-site inspection and maintenance costs.

[0003] In existing wastewater treatment equipment monitoring solutions, some technologies focus on directly controlling the final treatment effect. For example, by installing high-precision online water quality analysis instruments (such as chemical oxygen demand (COD) and ammonia nitrogen (NH3-N) sensors) at the equipment effluent, these solutions can achieve direct quantitative measurement of key pollutant indicators in the effluent, providing intuitive data support for determining whether the water quality meets standards. Other solutions focus on monitoring the basic operating conditions of the equipment. For example, by monitoring the operating current and frequency of aeration fans and water pumps, or conventional physical parameters such as dissolved oxygen (DO) concentration and pH value within the reaction unit, basic online sensing of the equipment's hardware start-up and shutdown status and macroscopic operating conditions is achieved.

[0004] However, the aforementioned existing technologies still have some limitations when applied to a large number of distributed sites. First, solutions relying on high-precision chemical sensors have high total cost of ownership due to their expensive equipment purchase costs and complex maintenance requirements (such as frequent probe cleaning, reagent replacement, and professional calibration), making them difficult to scale up. Simultaneously, this "results-oriented" monitoring has a significant lag; by the time water quality exceeds standards, the biochemical system within the reaction unit has often already experienced a serious malfunction, making it difficult to trace the root cause. Second, passive monitoring solutions relying on conventional physical or electrical parameters, while reducing costs, have inherent ambiguities in fault diagnosis. Wastewater treatment is a complex biochemical and physical coupled system; various faults with different causes (e.g., reduced microbial activity, aeration pipe blockage, or sudden increases in influent load) can easily lead to similar macroscopic parameter characteristics (e.g., decreased dissolved oxygen concentration). Relying solely on passively observed parameters, the monitoring system cannot effectively distinguish these fault types with similar characteristics, resulting in low diagnostic accuracy and failing to provide clear guidance for remote operation and maintenance. Summary of the Invention

[0005] To address the shortcomings of existing technologies, this invention provides a monitoring system for the operational status of decentralized sewage treatment equipment discharged into rivers. This system solves the problems of low diagnostic accuracy caused by the inability of passive monitoring schemes that rely on conventional physical parameters to distinguish between different faults such as reduced biological activity and physical blockage due to data ambiguity, while high-precision chemical sensing schemes are costly and complex to maintain.

[0006] To achieve the above objectives, the present invention provides the following technical solution:

[0007] This invention provides a system for monitoring the operational status of decentralized wastewater treatment equipment discharged into rivers, the system comprising:

[0008] The data acquisition module is used to collect raw acoustic signals and multi-point temperature data within the bioreactor unit of the wastewater treatment equipment.

[0009] A passive monitoring module, connected to the data acquisition module, is used to extract acoustic feature vectors and temperature feature vectors from the original acoustic signal and the multi-point temperature data, establish a static health baseline model, and calculate a static anomaly score based on the deviation of the acoustic feature vectors and temperature feature vectors from the static health baseline model.

[0010] An active detection and control module, connected to the passive monitoring module, is used to send a controlled disturbance command to the actuator of the wastewater treatment equipment when the static anomaly score exceeds a preset static anomaly threshold.

[0011] The dynamic response analysis module is used to collect the original acoustic signal and the multi-point temperature data during and after the execution of the controlled disturbance command, and extract dynamic response features from them;

[0012] The collaborative diagnosis module is used to integrate the static anomaly score and the dynamic response characteristics to generate system operation status diagnosis results.

[0013] Preferably, the data acquisition module includes:

[0014] An acoustic acquisition unit, including an underwater acoustic sensor, is used to acquire the raw acoustic signals;

[0015] The temperature acquisition unit includes temperature sensors, and the temperature sensor array is deployed at different locations within the bioreactor unit to collect the multi-point temperature data.

[0016] In one specific embodiment, the passive monitoring module includes:

[0017] An acoustic feature extraction unit is used to perform time-spectrum analysis on the original acoustic signal and extract the acoustic feature vector from the time-spectrum, wherein the acoustic feature vector includes the spectral centroid;

[0018] The temperature feature extraction unit is used to perform spatial interpolation based on the multi-point temperature data, construct a temperature field distribution model, and extract the temperature feature vector from the temperature field distribution model, wherein the temperature feature vector includes the average metabolic temperature rise.

[0019] Furthermore, the temperature feature extraction unit includes:

[0020] The inlet water temperature is obtained, and the difference between the average value of the multi-point temperature data in the bioreactor unit and the inlet water temperature is calculated as the average metabolic temperature rise.

[0021] Calculate the maximum temperature gradient or spatial thermodynamic variance in the temperature field distribution model and use it as a component of the temperature feature vector.

[0022] Preferably, the active detection control module includes:

[0023] The instruction generation unit is used to generate the controlled disturbance instruction when the static anomaly score exceeds the preset static anomaly threshold.

[0024] The instruction sending unit is used to send the controlled disturbance instruction to the actuator, wherein the actuator is the aeration blower or the inlet pump of the wastewater treatment equipment.

[0025] The controlled disturbance command is used to briefly change the aeration rate of the aeration blower or the inlet flow rate of the inlet pump within a preset duration.

[0026] Furthermore, the instruction generation unit is used to define the controlled disturbance instruction as a standardized pulse signal, the pulse signal including a preset disturbance amplitude and a preset duration.

[0027] In one specific embodiment, the dynamic response analysis module includes:

[0028] The thermal response analysis unit is used to calculate the average metabolic temperature rise based on the multi-point temperature data during and after the execution of the controlled disturbance command, and to extract thermal response features, including thermal response time lag and thermal response amplitude.

[0029] An acoustic response analysis unit is used to extract acoustic feature vectors and acoustic response features based on the original acoustic signal during and after the execution of the controlled disturbance command. The acoustic response features include acoustic recovery time.

[0030] Furthermore, the thermal response analysis unit is specifically used for:

[0031] The time required from the start of the execution of the controlled disturbance command to the peak value of the average metabolic temperature rise is calculated as the thermodynamic response time lag;

[0032] The maximum variation of the average metabolic temperature rise during the execution of the controlled disturbance command is calculated as the thermodynamic response amplitude.

[0033] Preferably, the collaborative diagnostic module includes:

[0034] A dynamic baseline establishment unit is used to establish a dynamic health baseline model, wherein the dynamic health baseline model is used to characterize the dynamic response characteristics to the controlled disturbance command under a healthy operating state;

[0035] The fusion diagnostic unit is used to calculate the deviation between the currently extracted dynamic response features and the dynamic health baseline model, and to fuse the deviation with the static anomaly score;

[0036] The diagnostic reasoning unit is used to calculate the posterior probability of different fault types based on the fused deviation and the static anomaly score, using a Bayesian network or expert rule base, and to take the fault type with the highest posterior probability as the diagnostic result of the system operating status.

[0037] In one specific embodiment, the dynamic baseline establishment unit specifically includes:

[0038] When the wastewater treatment equipment is in a healthy operating state, the controlled disturbance command is executed multiple times, and corresponding sets of dynamic response characteristics are collected.

[0039] Based on the multiple sets of dynamic response characteristics, their mean vector and covariance matrix are calculated to establish a multivariate normal distribution model characterizing the dynamic response characteristics, which serves as the dynamic health baseline model.

[0040] This invention provides a monitoring system for the operational status of decentralized wastewater treatment equipment discharged into rivers. It offers the following advantages:

[0041] 1. This invention acquires raw acoustic signals and multi-point temperature data within the bioreactor unit via a data acquisition module, and uses a passive monitoring module to extract acoustic and temperature feature vectors for state assessment. This approach replaces expensive and frequently calibrated chemical sensors (such as COD and ammonia nitrogen sensors) in traditional monitoring schemes with low-cost, high-stability physical sensors, reducing deployment costs and post-maintenance complexity. Simultaneously, it captures early physical signals reflecting biological activity, achieving low-cost, high-sensitivity monitoring of the operating status.

[0042] 2. When the passive monitoring module detects that the static anomaly score exceeds the limit, the present invention sends a controlled disturbance command to the actuator through the active detection control module, and the dynamic response analysis module extracts the dynamic response characteristics of the system under this disturbance. This closed-loop mechanism combining "passive monitoring and active detection" enables the system to obtain in-depth dynamic information that cannot be obtained by passive observation alone. This allows the collaborative diagnostic module to effectively distinguish fault types with similar static characteristics but different causes, such as distinguishing between reduced biological activity and physical faults such as blockage of aeration pipelines.

[0043] 3. The collaborative diagnostic module of this invention integrates static anomaly scores and dynamic response characteristics. Static anomaly scores reflect the degree to which the system's operating state deviates from the static health baseline, while dynamic response characteristics (such as thermal response lag and acoustic recovery time) characterize the inherent health and responsiveness under stress. By fusing these two different dimensions of information for diagnosis, compared to monitoring methods that rely solely on a single static feature or threshold, the accuracy and reliability of diagnostic results can be improved, effectively reducing the false alarm rate. Attached Figure Description

[0044] Figure 1 This is a system architecture diagram of the present invention.

[0045] Among them, 10 is the data acquisition module; 20 is the passive monitoring module; 30 is the active detection and control module; 40 is the dynamic response analysis module; and 50 is the collaborative diagnosis module. Detailed Implementation

[0046] The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0047] Reference Figure 1 , Figure 1 This is a structural block diagram of a decentralized wastewater treatment plant operation status monitoring system for discharge into a river, according to an embodiment of the present invention. The present invention provides an operation status monitoring system for decentralized wastewater treatment plant discharge into a river. The system is deployed at the bioreactor unit of the decentralized wastewater treatment plant and communicates with the control system (e.g., a programmable logic controller, PLC) of the plant. In one specific embodiment, the system includes:

[0048] The data acquisition module 10 is deployed in the bioreactor unit of the wastewater treatment equipment and is used to collect raw acoustic signals and multi-point temperature data in the bioreactor unit.

[0049] The passive monitoring module 20 is connected to the data acquisition module 10. The passive monitoring module 20 is used to process the received raw acoustic signal and multi-point temperature data in real time during normal system operation, extracting acoustic feature vectors and temperature feature vectors from them.

[0050] The passive monitoring module 20 is also used to establish a static health baseline model based on the acoustic feature vector and the temperature feature vector collected under healthy operating conditions.

[0051] During the monitoring process, the passive monitoring module 20 calculates the static anomaly score based on the deviation of the currently extracted acoustic feature vector and temperature feature vector from the static health baseline model.

[0052] An active detection control module 30 is connected to the passive monitoring module 20. The active detection control module 30 is used to trigger an active detection process when the static anomaly score calculated by the passive monitoring module 20 exceeds a preset static anomaly threshold. After triggering, the active detection control module 30 sends a preset controlled disturbance command to the actuator of the wastewater treatment equipment (e.g., the controller of the aeration blower or the influent pump).

[0053] The dynamic response analysis module 40 acquires the original acoustic signal and the multi-point temperature data from the data acquisition module 10 during and after the active detection control module 30 sends the controlled disturbance command.

[0054] The dynamic response analysis module 40 is used to analyze the transient changes of the data under controlled disturbances and extract a set of dynamic response features. The dynamic response features characterize the system's response to the controlled disturbance command.

[0055] The collaborative diagnosis module 50 is connected to both the passive monitoring module 20 and the dynamic response analysis module 40. The collaborative diagnosis module 50 is used to fuse the static anomaly score calculated by the passive monitoring module 20 and the dynamic response features extracted by the dynamic response analysis module 40.

[0056] Based on the fused information, the collaborative diagnostic module 50 comprehensively evaluates the operating status of the wastewater treatment equipment and generates system operating status diagnostic results, such as indicating specific fault types.

[0057] In the above system architecture, the data acquisition module 10 acts as the perception layer, continuously acquiring physical signals. The passive monitoring module 20 acts as a routine monitoring unit, performing baseline comparison and anomaly identification. The active detection control module 30 acts as a state switching and control unit, initiating the active detection process upon receiving an anomaly trigger signal. The dynamic response analysis module 40 acts as a deep feature extraction unit, used to acquire transient response data of the system under stress. The collaborative diagnosis module 50 acts as a decision-making unit, fusing static and dynamic information to make a final diagnosis.

[0058] See attached document Figure 1 The data acquisition module 10 is the physical sensing entry point of the monitoring system. It is deployed within the bioreactor unit of the wastewater treatment equipment or its auxiliary pipelines to acquire raw physical signals reflecting the operating status of the bioreactor unit in real time. Specifically, the data acquisition module 10 includes an acoustic acquisition unit and a temperature acquisition unit.

[0059] In one embodiment, the acoustic acquisition unit includes at least one underwater acoustic sensor, such as a broadband hydrophone. The underwater acoustic sensor is submerged in the mixture of the bioreactor. Its installation location is chosen to represent the average acoustic environment of the reactor and avoid direct interference from high-intensity noise sources; for example, it is installed in the middle of the reactor wall, away from the direct outlet of the aeration pipe or the impeller blades of the agitator, to prevent aeration bubbles from directly impacting the sensor probe or excessive mechanical vibration signals from causing signal saturation.

[0060] The acoustic acquisition unit also includes an analog-to-digital converter (ADC). The analog acoustic signals captured by the underwater acoustic sensor are sent to the ADC for sampling at a preset frequency. Sampling is performed. To fully capture broadband information during biological metabolism and aeration processes, the sampling frequency... The frequency is set to be no lower than 44.1 kHz. The output of the acoustic acquisition unit is a digitized, time-domain raw acoustic signal. ,in This is the index for the sampling points.

[0061] The temperature acquisition unit includes A temperature sensor deployed inside the bioreactor unit and a temperature sensor deployed at the inlet of the bioreactor unit.

[0062] To accurately capture the slight temperature rise caused by microbial metabolic activity, the The internal temperature sensor is a high-precision temperature sensor, such as a PT1000 platinum resistance temperature detector (RTD) with a resolution better than 0.01℃ or a high-precision thermistor.

[0063] The An internal temperature sensor is deployed in an array within the bioreactor unit. In one specific embodiment, the array is deployed in a three-dimensional mesh layout, meaning the sensors are fixed at different depths within the bioreactor unit. ) and different horizontal coordinates ( , On top. Each internal temperature sensor ( Precise three-dimensional spatial coordinates It is pre-calibrated and stored in the system as the basis for subsequent construction of temperature field distribution models.

[0064] The output of the temperature acquisition unit is a set of multi-point temperature data, including the... Temperature data collected by an internal sensor and the inlet water temperature collected by the inlet water temperature sensor. The data acquisition module 10 ensures that the acquisition of the raw acoustic signal and the multi-point temperature data is synchronized in time.

[0065] Reference Figure 1 The passive monitoring module 20 is connected to the data acquisition module 10 and is used to continuously process and analyze the collected data under the normal operating conditions of the sewage treatment equipment.

[0066] In one embodiment, the passive monitoring module 20 includes an acoustic feature extraction unit. The acoustic feature extraction unit receives the raw acoustic signal output by the data acquisition module 10. First, the unit... Perform signal preprocessing, such as applying band-stop filters to filter out deterministic noise at specific frequencies (e.g., 50Hz power frequency) generated by the power supply system or mechanical pump body.

[0067] After preprocessing, the acoustic feature extraction unit performs time-spectrum analysis on the signal using short-time Fourier transform (STFT) to obtain a time-spectrum diagram. ,in For frequency index, Indexing for time frames. Then, from each time frame of the time spectrogram... Extract the acoustic feature vector .

[0068] The acoustic feature vector includes the spectral centroid. Spectral centroid It reflects the center position of the signal power spectrum, and its specific calculation method is as follows:

[0069] ;

[0070] In the formula, In time frame Frequency Index The power; The centroid of the spectrum; This is a frequency index. Furthermore, the acoustic feature vector... It can also include spectral entropy or sub-band energy of a specific frequency band.

[0071] The passive monitoring module 20 also includes a temperature feature extraction unit. The temperature feature extraction unit receives multi-point temperature data output by the data acquisition module 10. Inlet water temperature and pre-stored Spatial coordinates of an internal sensor .

[0072] The temperature feature extraction unit utilizes the and A continuous three-dimensional temperature field distribution model is constructed using spatial interpolation algorithms (such as radial basis function interpolation or kriging interpolation). ,in , where is any spatial point within the bioreaction unit.

[0073] The temperature feature extraction unit extracts the temperature feature vector from the temperature field distribution model and the original temperature data. The temperature feature vector includes the average metabolic temperature rise. The average metabolic temperature rise The calculation method is as follows:

[0074] ;

[0075] In the formula, This represents the total number of internal temperature sensors. This represents the average metabolic temperature rise. This refers to the inlet water temperature. Indicates the first An internal temperature sensor in time The collected temperature.

[0076] The temperature feature vector It also includes the temperature field distribution model. Maximum temperature gradient , or the aforementioned Spatial thermal variance of internal temperature sensor readings The maximum temperature gradient is used to characterize the region of strongest metabolic exothermic activity, while the spatial thermodynamic variance is used to characterize the uniformity of temperature distribution.

[0077] The passive monitoring module 20 is also used to establish a static health baseline model. It also calculates static anomaly scores. When the wastewater treatment equipment is in a healthy operating state, the module collects a large number of acoustic feature vectors. and temperature eigenvectors And combine them into a static state vector. .

[0078] The module employs a probability density estimation algorithm, such as a Gaussian mixture model (GMM), to train the collected static health state vector sample set, thereby establishing the static health baseline model. The model It represents the probability distribution of health status.

[0079] During routine monitoring, the passive monitoring module 20 calculates the current static state vector in real time. The static anomaly score Calculated as current In the static health baseline model Negative log-likelihood value:

[0080] ;

[0081] In the formula, This refers to static anomaly scores; This is the static state vector; This is a static health baseline model; It is a logarithmic function; It is by Defined probability density function; The value and The degree of deviation from a healthy state is positively correlated. It will be sent to the active detection control module 30.

[0082] Reference Figure 1 The active detection control module 30 is connected to the passive monitoring module 20 and establishes a communication connection with the actuator of the sewage treatment equipment (or its upstream PLC controller).

[0083] The active detection control module 30 is used to receive the static anomaly score calculated in real time by the passive monitoring module 20. The active detection and control module 30 internally stores a preset static anomaly threshold. The static anomaly threshold is determined based on the statistical distribution of the static anomaly scores of the wastewater treatment equipment under healthy operating conditions.

[0084] In one embodiment, the active detection control module 30 includes an instruction generation unit. The instruction generation unit, within a monitoring cycle, will receive... and A comparison is performed. When the instruction generation unit determines... When the time comes, the active detection process is triggered, and the controlled disturbance command is generated.

[0085] The controlled disturbance command is defined as a standardized pulse signal. This controlled disturbance command is used to adjust the control setpoint of the actuator. At its current baseline value Based on this, a brief change is made according to a preset perturbation waveform. In one specific implementation, this waveform is a rectangular pulse, mathematically defined as:

[0086] ;

[0087] In the formula, It is the start time of the disturbance; It is a standard rectangular window function; The preset disturbance amplitude for the controlled disturbance command; To control the set value; Baseline value; The duration of the increased aeration frequency; It is a time variable.

[0088] The It is the preset disturbance amplitude of the controlled disturbance command, and the preset disturbance amplitude of the controlled disturbance command is the preset duration of the command. and The value is pre-calibrated and stored in the instruction generation unit to ensure that the perturbation applied by each active probe is standardized.

[0089] The actuator is either an aeration blower or an influent pump of the wastewater treatment equipment. If the actuator is an aeration blower, then... Corresponding to the current aeration fan frequency, Corresponding to a preset frequency increment, This corresponds to the duration of the increased aeration frequency. If the actuator is an inlet pump, then... Corresponding to the current inlet pump frequency or rated flow rate, This corresponds to a preset traffic increment.

[0090] The active detection control module 30 further includes a command sending unit. The command sending unit is used to transmit the controlled disturbance command generated by the command generation unit (i.e., in...) The setting value will be adjusted at any time. and in Restored to The control logic is sent to the programmable logic controller (PLC) or frequency converter of the actuator via an industrial communication protocol.

[0091] While the instruction sending unit sends the controlled disturbance instruction, the active detection and control module 30 also sends a synchronization trigger signal to the dynamic response analysis module 40 so that it can start high-frequency acquisition and analysis of dynamic response data.

[0092] Reference Figure 1 The dynamic response analysis module 40 is connected to the active detection control module 30 and the data acquisition module 10, respectively. The dynamic response analysis module 40 is used to respond to a synchronization trigger signal (indicating that the controlled disturbance command is in effect) sent by the active detection control module 30. It is activated when execution begins.

[0093] In one embodiment, the dynamic response analysis module 40 includes a thermal response analysis unit. The thermal response analysis unit performs analysis within a preset time window during and after the execution of the controlled disturbance command (e.g., from...). to The system frequently acquires the multi-point temperature data and the inlet water temperature from the data acquisition module 10.

[0094] The thermal response analysis unit calculates the average metabolic temperature rise in real time based on the multi-point temperature data and the inlet water temperature. The calculation method is consistent with the method used by the temperature feature extraction unit in the passive monitoring module 20.

[0095] The thermal response analysis unit is used to analyze the mean metabolic temperature rise. The thermal response features are extracted from the time series curves.

[0096] The thermal response analysis unit first, within the analysis window Within, search the mean metabolic temperature rise peak and the time of its occurrence ,in .

[0097] The aforementioned thermodynamic response characteristics include thermodynamic response time delay. The thermodynamic response time delay Calculated to be executed from the start of the controlled disturbance command Until the average metabolic temperature rise reaches its peak Time required:

[0098] ;

[0099] In the formula, This is due to the time lag in the thermal response; The average metabolic temperature rise reaches its peak; It is the start time of the disturbance.

[0100] The thermal response characteristics also include the thermal response amplitude. The thermal response amplitude The mean metabolic temperature rise is calculated as the maximum variation during the execution of the controlled perturbation command, i.e., the peak value. Reference value at the start of the disturbance The difference between them:

[0101] ;

[0102] In the formula, The amplitude of the thermal response; Peak moment The corresponding average metabolic temperature rise; This is the baseline value at the start of the disturbance.

[0103] In one embodiment, the dynamic response analysis module 40 further includes an acoustic response analysis unit. The acoustic response analysis unit acquires the raw acoustic signal from the data acquisition module 10 within the analysis time window. .

[0104] The acoustic response analysis unit uses the same method as the acoustic feature extraction unit in the passive monitoring module 20 to calculate the acoustic feature vector in real time. .

[0105] The acoustic response analysis unit is used to extract acoustic response features, including acoustic recovery time. .

[0106] The acoustic response analysis unit first bases its analysis on the real-time calculations. The static health baseline model established with the passive monitoring module 20 (or only the portion containing acoustic features), calculate the acoustic anomaly score. .

[0107] The controlled disturbance command is in End of time (where) (The acoustic recovery time is a preset duration). Calculated as, from the end time of the controlled disturbance command Start, until the acoustic anomaly score The acoustic recovery threshold was first lowered. The following are the times that have elapsed. :

[0108] ;

[0109] In the formula, Acoustic recovery time; The preset acoustic recovery threshold; It is the start time of the disturbance; The duration of the increased aeration frequency; This represents a variable that indicates the time elapsed since the end of the disturbance. To find the minimum value in the set that satisfies the subsequent conditions, i.e., to find the minimum time that meets the conditions. ; The acoustic anomaly score.

[0110] Finally, the dynamic response analysis module 40 extracts the thermodynamic response features. and the acoustic response characteristics ( This is combined into a dynamic response feature vector. The vector is then sent to the collaborative diagnostic module 50.

[0111] Reference Figure 1 The collaborative diagnostic module 50 is connected to both the passive monitoring module 20 and the dynamic response analysis module 40. The collaborative diagnostic module 50 is used to collect static monitoring information and dynamic detection information of the system to make a final system operation status diagnosis result.

[0112] In one embodiment, the collaborative diagnostic module 50 includes a dynamic baseline establishment unit, a fusion diagnostic unit, and a diagnostic inference unit.

[0113] The dynamic baseline establishment unit is used to establish the dynamic health baseline model during the phase when the wastewater treatment equipment is confirmed to be in a healthy operating state (e.g., after system initialization or maintenance calibration). .

[0114] During this healthy operating phase, the dynamic baseline establishment unit controls the active detection control module 30 to repeatedly execute... The standardized controlled disturbance command described below. For the first Second-rate Execution: The unit collects the corresponding data from the dynamic response analysis module 40. Group dynamic response feature vector .

[0115] After collecting After generating the dynamic response feature vector, the dynamic baseline establishment unit calculates the mean vector of this set of health status samples. Covariance Matrix

[0116] ;

[0117] ;

[0118] In the formula, This is the mean vector of the health status samples; It is the covariance matrix; For the first Group dynamic response feature vector; The total number of sets of dynamic response feature vectors; For superscript, the transpose operator in mathematics.

[0119] The dynamic baseline establishment unit will calculate the... and The parameters of the multivariate normal distribution model characterizing the dynamic response features are stored. This multivariate normal distribution model serves as the dynamic health baseline model. .

[0120] The fusion diagnostic unit performs fusion calculations after completing an active detection process during routine system monitoring. The fusion diagnostic unit receives the static anomaly score that triggered the detection from the passive monitoring module 20. and receive the currently extracted dynamic response feature vector from the dynamic response analysis module 40. .

[0121] The fusion diagnostic unit first calculates the current... With the dynamic health baseline model The deviation. In a specific implementation, this deviation... For the Compared to and The Mahalanobis distance of the multivariate normal distribution is defined.

[0122] The fusion diagnostic unit integrates this deviation. and the static anomaly score For example, combining them into a fused feature vector. The fused feature vector is then sent to the diagnostic inference unit.

[0123] The diagnostic reasoning unit is used to base its analysis on the received fusion feature vector. The posterior probability of different failure types is calculated through inference. These failure types are predefined, and include, for example, reduced biological activity, aeration blockage, or sludge bulking.

[0124] In one embodiment, the diagnostic inference unit employs a pre-trained Bayesian network. This Bayesian network uses the fused feature vector... The amount and The evidence node is defined as the default fault type, and the query node is defined as the preset fault type. The diagnostic reasoning unit calculates the fault type by performing probabilistic reasoning on this Bayesian network. posterior probability .

[0125] In another embodiment, the diagnostic reasoning unit employs an expert rule base. This rule base contains a set of logical rules that will... Different regions in the vector space are mapped to specific fault types.

[0126] Finally, the diagnostic reasoning unit will determine the fault type with the highest posterior probability (i.e., The fault type, or the fault type triggered by the expert rule base, is output as the system operation status diagnosis result.

Claims

1. A monitoring system for the operational status of decentralized sewage treatment equipment discharged into rivers, characterized in that, The system includes: The data acquisition module is used to collect raw acoustic signals and multi-point temperature data within the bioreactor unit of the wastewater treatment equipment. A passive monitoring module, connected to the data acquisition module, is used to extract acoustic feature vectors and temperature feature vectors from the original acoustic signal and the multi-point temperature data, establish a static health baseline model, and calculate a static anomaly score based on the deviation of the acoustic feature vectors and temperature feature vectors from the static health baseline model. An active detection and control module, connected to the passive monitoring module, is used to send a controlled disturbance command to the actuator of the wastewater treatment equipment when the static anomaly score exceeds a preset static anomaly threshold. The dynamic response analysis module is used to collect the original acoustic signal and the multi-point temperature data during and after the execution of the controlled disturbance command, and extract dynamic response features from them; The collaborative diagnosis module is used to fuse the static anomaly score and the dynamic response characteristics to generate system operation status diagnosis results; The passive monitoring module includes: An acoustic feature extraction unit is used to perform time-spectrum analysis on the original acoustic signal and extract the acoustic feature vector from the time-spectrum, wherein the acoustic feature vector includes the spectral centroid; The temperature feature extraction unit is used to perform spatial interpolation based on the multi-point temperature data, construct a temperature field distribution model, and extract the temperature feature vector from the temperature field distribution model, wherein the temperature feature vector includes the average metabolic temperature rise. The active detection control module includes: The instruction generation unit is used to generate the controlled disturbance instruction when the static anomaly score exceeds the preset static anomaly threshold. The instruction sending unit is used to send the controlled disturbance instruction to the actuator, wherein the actuator is the aeration blower or the inlet pump of the wastewater treatment equipment. The controlled disturbance command is used to briefly change the aeration volume of the aeration blower or the inlet flow rate of the inlet pump within a preset duration. The dynamic response analysis module includes: The thermal response analysis unit is used to calculate the average metabolic temperature rise based on the multi-point temperature data during and after the execution of the controlled disturbance command, and to extract thermal response features, including thermal response time lag and thermal response amplitude. An acoustic response analysis unit is used to extract acoustic feature vectors and acoustic response features based on the original acoustic signal during and after the execution of the controlled disturbance command. The acoustic response features include acoustic recovery time.

2. The decentralized sewage treatment equipment operation status monitoring system for discharge into rivers according to claim 1, characterized in that, The data acquisition module includes: The acoustic acquisition unit includes an underwater acoustic sensor for acquiring raw acoustic signals; The temperature acquisition unit includes temperature sensors, and the temperature sensor array is deployed at different locations within the bioreactor unit to collect the multi-point temperature data.

3. The system for monitoring the operational status of decentralized sewage treatment equipment discharged into rivers according to claim 1, characterized in that, The temperature feature extraction unit includes: The inlet water temperature is obtained, and the difference between the average value of the multi-point temperature data in the bioreactor unit and the inlet water temperature is calculated as the average metabolic temperature rise. Calculate the maximum temperature gradient or spatial thermodynamic variance in the temperature field distribution model and use it as a component of the temperature feature vector.

4. The system for monitoring the operational status of decentralized sewage treatment equipment discharged into rivers according to claim 1, characterized in that, The instruction generation unit is used to define the controlled disturbance instruction as a standardized pulse signal, the pulse signal including a preset disturbance amplitude and a preset duration.

5. The system for monitoring the operational status of decentralized sewage treatment equipment discharged into rivers according to claim 1, characterized in that, The thermodynamic response analysis unit is specifically used for: The time required from the start of the execution of the controlled disturbance command to the peak value of the average metabolic temperature rise is calculated as the thermodynamic response time lag; The maximum variation of the average metabolic temperature rise during the execution of the controlled disturbance command is calculated as the thermodynamic response amplitude.

6. The system for monitoring the operational status of decentralized sewage treatment equipment discharged into rivers according to claim 1, characterized in that, The collaborative diagnostic module includes: A dynamic baseline establishment unit is used to establish a dynamic health baseline model, wherein the dynamic health baseline model is used to characterize the dynamic response characteristics to the controlled disturbance command under a healthy operating state; The fusion diagnostic unit is used to calculate the deviation between the currently extracted dynamic response features and the dynamic health baseline model, and to fuse the deviation with the static anomaly score; The diagnostic reasoning unit is used to calculate the posterior probability of different fault types based on the fused deviation and the static anomaly score, using a Bayesian network or expert rule base, and to take the fault type with the highest posterior probability as the diagnostic result of the system operating status.

7. The system for monitoring the operational status of decentralized sewage treatment equipment discharged into rivers according to claim 6, characterized in that, The dynamic baseline establishment unit specifically includes: When the wastewater treatment equipment is in a healthy operating state, the controlled disturbance command is executed multiple times, and corresponding sets of dynamic response characteristics are collected. Based on the multiple sets of dynamic response characteristics, their mean vector and covariance matrix are calculated to establish a multivariate normal distribution model characterizing the dynamic response characteristics, which serves as the dynamic health baseline model.

Images