Remote sewage treatment system based on big data model

Abstract

The invention discloses a remote sewage treatment system based on a big data model. The remote sewage treatment system comprises an information collection module (100), a data processing module (200), a data prediction module (300) and a visualization module (400), the information collection module (100) comprises a sensing network unit (101) and an edge signal calculation unit (102), the sensing network unit (101) comprises a sensor arranged at the position of a sewage treatment facility, and the edge signal calculation unit (102) is used for preprocessing collected data; the data processing module (200) verifies the causal relationship among the water quality parameters by adopting a convergent cross-mapping algorithm; and the data prediction module (300) is used for establishing a bidirectional neural network and predicting an expected water quality index. The method has the advantages that the method has better effects in the aspects of data acquisition density, feature interpretation, prediction precision, operation and maintenance efficiency and the like compared with a traditional system, and especially under the dynamic working condition, the error of the bidirectional neural network model is controlled at a low numerical value and is obviously improved compared with a traditional method.

Description

Technical Field

[0001] This invention relates to the field of factory wastewater treatment, and in particular to a remote wastewater treatment system based on a big data model. Background Technology

[0002] With the acceleration of urbanization, the intelligent upgrading of wastewater treatment systems has become an important issue in the environmental protection field. Traditional wastewater treatment systems mostly adopt a centralized monitoring architecture based on SCADA, and its technical limitations are mainly reflected in: 1. Data collection dimension is singular. Existing systems typically deploy single-point sensors in key process sections and use linear interpolation to estimate the overall water quality distribution of the plant. This fails to effectively capture the three-dimensional gradient changes of dissolved oxygen in the biological treatment tank and the spatial heterogeneity of sludge concentration, resulting in deviations between the control strategy and actual operating conditions.

[0003] 2. The model's prediction accuracy is limited. Mainstream prediction methods often employ ARIMA or single-layer LSTM models, failing to adequately consider the nonlinear causal relationships between water quality parameters. For instance, the impact of aeration rate adjustments on ammonia nitrogen removal exhibits a time lag effect, making it difficult for traditional time-series models to accurately model such cross-parameter dynamic coupling relationships.

[0004] 3. Low efficiency of edge-cloud collaboration Existing solutions typically upload raw data directly to the cloud for processing without performing feature pre-extraction at the edge, resulting in high bandwidth usage and long cloud model update cycles, making it difficult to respond promptly to sudden changes in influent water quality.

[0005] 4. The contradiction between data privacy and model generalization When optimizing multiple plant areas collaboratively, existing federated learning solutions often use simple parameter averaging, without designing privacy protection mechanisms for wastewater treatment scenarios, which poses a risk of process parameter leakage. At the same time, they ignore the impact of differences in the aging of equipment in different plant areas on model migration.

[0006] 5. The disconnect between visualization and control Traditional 3D visualization systems only display data statically and do not map the prediction results to the generation of control commands in real time. Operators still need to rely on experience to manually adjust equipment parameters, and the response delay is significant in emergency situations such as rainstorms. Summary of the Invention

[0007] The purpose of this section is to outline some aspects of embodiments of the present invention and to briefly describe some preferred embodiments. Simplifications or omissions may be made in this section, as well as in the abstract and title of this application, to avoid obscuring the purpose of these documents; however, such simplifications or omissions should not be construed as limiting the scope of the invention.

[0008] In view of the problems existing in the above and / or existing remote sewage treatment systems based on big data models, this invention is proposed.

[0009] Therefore, the problem to be solved by the present invention is how to provide a solution that enables remote wastewater treatment.

[0010] To address the aforementioned technical problems, this invention provides the following technical solution: a remote wastewater treatment system based on a big data model, comprising, It includes an information collection module, a data processing module, a data prediction module, and a visualization module; The information collection module includes a sensor network unit and an edge signal computing unit. The sensor network unit includes sensors deployed at the location of the wastewater treatment facility. The edge signal computing unit is used to preprocess the data collected by the sensor network unit. The data processing module includes a hybrid storage unit and a feature engineering unit. The hybrid storage unit includes a time-series database and a relational database. The feature engineering unit uses a convergent cross-mapping algorithm to verify the causal relationship between water quality parameters. The data prediction module is used to establish a bidirectional neural network, collect data from the edge signal calculation unit after preprocessing, and predict expected water quality indicators. The visualization module collects the output data from the hybrid storage unit and the data prediction module and displays it in the form of images.

[0011] As a preferred embodiment of the remote wastewater treatment system based on a big data model according to the present invention, the sensor network unit includes sensors deployed at the location of the wastewater treatment facility, including: The sensor network unit deploys sensors at the inlet, biological treatment tank, sedimentation tank, and outlet of the wastewater treatment facility; A three-dimensional point layout is adopted in the biochemical reaction tank, with at least three detection layers set along the vertical direction, and composite sensors distributed in a honeycomb grid in each layer; An ultraviolet absorption spectroscopy detection component is installed at the water outlet.

[0012] As a preferred embodiment of the remote sewage treatment system based on a big data model according to the present invention, the edge signal calculation unit is used to preprocess the data collected by the sensor network unit, including: Edge-cloud data collaborative control equation: ; Where ε(t) represents the cooperative effect, t is the synchronization timestamp, ∂Ω is the closed path formed by the sensor network in the complex plane, and F edge (z) is the feature function of the preprocessed data of the edge node z, Gcloud Let Re(⋅) be the analytic extension function of the corresponding node in the cloud, where Re(⋅) takes the real part of the complex function, Li2(⋅) is the double logarithm function, and ρ sync Let ζ be the data synchronization rate between the edge and the cloud, ζ be the network jitter coefficient, Si(⋅) be the sine integral function, and δ be the data synchronization rate between the edge and the cloud. latency This represents the transmission delay in milliseconds.

[0013] As a preferred embodiment of the remote wastewater treatment system based on a big data model according to the present invention, the data prediction module is used to establish a bidirectional neural network, collect data preprocessed from the edge signal computing unit, and predict expected water quality indicators, including: Water quality prediction-control coupling equations: ; Where P(t) represents the system control sensitivity, Proj g (⋅) is the projection operator on the Lie algebra g, ∇ θ L BiLSTM Let J be the gradient of the loss function of the bidirectional neural network, so(3) be the three-dimensional rotation group norm, and J be the gradient of the loss function. CCM T Let Det(⋅) be the convergent cross-mapping matrix, and Σ be the determinant. sensor For the sensor data covariance matrix, ent KL ν is the divergence entropy between the predicted and measured values, and ν is the equipment aging correction factor.

[0014] As a preferred embodiment of the remote wastewater treatment system based on a big data model described in this invention, the hybrid storage unit includes: Time-series databases, relational databases, and distributed file systems are used for hierarchical storage of real-time water quality data, equipment metadata, and unstructured data.

[0015] As a preferred embodiment of the remote wastewater treatment system based on a big data model according to the present invention, the visualization module collects the output data of the hybrid storage unit and the data prediction module and displays it in image form, including: The visualization module outputs parameters to the PLC controller, and an alarm threshold is set. When the wastewater quality value exceeds the wastewater quality value predicted by the data prediction module, an email notification is triggered.

[0016] As a preferred embodiment of the remote wastewater treatment system based on a big data model described in this invention, the time-series database, relational database, and distributed file system include: Use a time-series database to store real-time streaming data from sensors; Relational databases store device metadata and process parameters; The distributed file system stores video surveillance and equipment maintenance log data.

[0017] As a preferred embodiment of the remote wastewater treatment system based on a big data model described in this invention, the feature engineering unit uses a convergent cross-mapping algorithm to verify the causal relationship between water quality parameters, including: The feature engineering unit calculates the spatial gradient of adjacent sensor data, extracts time-series statistical features through a sliding window, and uses a convergent cross-mapping algorithm to verify the causal relationship between water quality parameters. The time-series statistical features include mean, variance, trend slope, and permutation entropy.

[0018] In a second aspect, some embodiments of the present invention provide an electronic device, including: one or more processors; and a storage device having one or more programs stored thereon, wherein when the one or more programs are executed by the one or more processors, the one or more processors implement the method described in any of the implementations of the first aspect above.

[0019] Thirdly, some embodiments of the present invention provide a computer-readable medium having a computer program stored thereon, wherein the program, when executed by a processor, implements the method described in any of the implementations of the first aspect above.

[0020] The beneficial effects of this invention are that it proposes a remote sewage treatment system based on a big data model. This system has better performance than traditional systems in terms of data acquisition density, feature interpretability, prediction accuracy and operation and maintenance efficiency. In particular, under dynamic operating conditions, its bidirectional neural network model error is controlled at a low value, which is a significant improvement over traditional methods. Attached Figure Description

[0021] To more clearly illustrate the technical solutions in this invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0022] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort. Wherein: Figure 1 This is a flow chart of a remote wastewater treatment system based on a big data model in Example 1; Figure 2This is a schematic diagram illustrating the monitoring effect of a remote sewage treatment system based on a big data model in Example 1. Detailed Implementation

[0023] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0024] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and those skilled in the art can make similar extensions without departing from the spirit of the invention. Therefore, the invention is not limited to the specific embodiments disclosed below.

[0025] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The phrase "in one embodiment" appearing in different places in this specification does not necessarily refer to the same embodiment, nor is it a single or selective embodiment that is mutually exclusive with other embodiments. Example 1

[0026] Reference Figure 1 and Figure 2 This is the first embodiment of the present invention, which provides a remote wastewater treatment system based on a big data model, comprising: It includes an information collection module 100, a data processing module 200, a data prediction module 300, and a visualization module 400; The information collection module 100 includes a sensor network unit 101 and an edge signal calculation unit 102. The sensor network unit 101 includes sensors deployed at the location of the sewage treatment facility. The edge signal calculation unit 102 is used to preprocess the data collected by the sensor network unit 101. The sensor network unit 101 includes sensors deployed at the location of the wastewater treatment facility, including: The sensor network unit 101 deploys sensors at the inlet, biological treatment tank, sedimentation tank and outlet of the wastewater treatment facility; A three-dimensional point layout is adopted in the biochemical reaction tank, with at least three detection layers set along the vertical direction, and composite sensors distributed in a honeycomb grid in each layer; An ultraviolet absorption spectroscopy detection component is installed at the water outlet. A three-dimensional distribution of dissolved oxygen sensors is used in the biochemical reaction tank, including 0.5m below the water surface, the tank bottom, and the intermediate layer, with one set of dissolved oxygen composite sensors configured for every 100m³. An ultrasonic sludge concentration meter (range 0-20g / L, accuracy ±0.5%) is installed on the sludge return pipeline. An ultraviolet absorption spectrometer (wavelength 254nm) is deployed at the outlet of the disinfection pool to detect residual chlorine in real time.

[0027] The edge signal calculation unit 102 is used to preprocess the data collected by the sensor network unit 101, including: Edge-cloud data collaborative control equation: ; Where ε(t) represents the cooperative effect, t is the synchronization timestamp, ∂Ω is the closed path formed by the sensor network in the complex plane, and F edge (z) is the feature function of the preprocessed data of the edge node z, G cloud Let Re(⋅) be the analytic extension function of the corresponding node in the cloud, where Re(⋅) takes the real part of the complex function, Li2(⋅) is the double logarithm function, and ρ sync Let ζ be the data synchronization rate between the edge and the cloud, ζ be the network jitter coefficient, Si(⋅) be the sine integral function, and δ be the data synchronization rate between the edge and the cloud. latency ε(t) represents the transmission delay in milliseconds. ε(t) ∈ (−∞, +∞), where a positive value indicates good edge-cloud data coordination, and a negative value indicates that a retransmission mechanism needs to be triggered. When ε(t) > 1.5, the compression algorithm is automatically optimized; when ε(t) < −0.8, the system switches to a 4G redundant channel.

[0028] The data processing module 200 includes a hybrid storage unit 201 and a feature engineering unit 202. The hybrid storage unit 201 includes a time-series database and a relational database. The feature engineering unit 202 uses a convergent cross-mapping algorithm to verify the causal relationship between water quality parameters. The hybrid storage unit 201 includes: Time-series databases, relational databases, and distributed file systems are used for hierarchical storage of real-time water quality data, equipment metadata, and unstructured data.

[0029] Time-series databases, relational databases, and distributed file systems include: Use a time-series database to store real-time streaming data from sensors; Relational databases store device metadata and process parameters; A distributed file system stores video surveillance and equipment maintenance log data. Feature engineering unit 202 uses a convergent cross-mapping algorithm to verify the causal relationships between water quality parameters, including: The feature engineering unit 202 calculates the spatial gradient of adjacent sensor data, extracts time-series statistical features through a sliding window, and uses a convergent cross-mapping algorithm to verify the causal relationship between water quality parameters. The time-series statistical features include mean, variance, trend slope, and permutation entropy.

[0030] The data prediction module 300 is used to establish a bidirectional neural network, collect preprocessed data from the edge signal calculation unit 102, and predict expected water quality indicators. The data prediction module 300's function of establishing a bidirectional neural network, collecting preprocessed data from the edge signal calculation unit 102, and predicting expected water quality indicators includes: Water quality prediction-control coupling equations: ; Where P(t) represents the system control sensitivity, Proj g (⋅) is the projection operator on the Lie algebra g, ∇ θ L BiLSTM Let J be the gradient of the loss function of the bidirectional neural network, so(3) be the three-dimensional rotation group norm, and J be the gradient of the loss function. CCM T Let Det(⋅) be the convergent cross-mapping matrix, and Σ be the determinant. sensor For the sensor data covariance matrix, ent KL Let ν be the divergence entropy between the predicted and measured values, and ν be the equipment aging correction factor. P(t) ∈ [0, π / 2), and a larger value indicates higher system control sensitivity. When P(t) > 1.2, the high-frequency sampling mode is activated, and when P(t) < 0.3, the sensor calibration procedure is triggered.

[0031] The visualization module 400 collects the output data of the hybrid storage unit 201 and the data prediction module 300 and displays it in the form of images. Example 2

[0032] The second embodiment of the present invention differs from the first embodiment in that it further includes an experimental preparation and implementation process: I. Experiment Preparation and Implementation Process This experiment selected a wastewater treatment plant in an industrial park of a certain city as the test scenario. The plant has a daily treatment capacity of approximately 100,000 tons and includes core process sections such as biological treatment, sedimentation tanks, and disinfection tanks. The test objective was to verify the system's performance in real-time monitoring, data prediction, and visualization, and to compare it with a traditional SCADA system. The experiment lasted 30 days and was divided into four phases: sensor deployment, data acquisition, model training, and prediction verification.

[0033] 1. Sensor Network Deployment The sensor network unit 101 deploys 42 multi-parameter water quality sensors, covering the inlet, aeration tank, secondary sedimentation tank, and outlet. Monitored parameters include pH, dissolved oxygen (DO), chemical oxygen demand (COD), ammonia nitrogen (NH3-N), total phosphorus (TP), and suspended solids (SS), with a sampling frequency of once per minute. The sensors employ a hybrid network of RS485 bus and LoRa wireless to ensure signal stability in high-noise environments. The edge signal computing unit 102 uses an embedded ARM processor and deploys data cleaning algorithms: median filtering is applied to outliers (such as instantaneous DO fluctuations >2 mg / L), and time-series interpolation is used to fill in missing data. The preprocessed data compression rate is increased to 35% of the original data.

[0034] 2. Hybrid Storage and Feature Engineering Hybrid storage unit 201 uses the time-series database InfluxDB to store raw sensor data (approximately 12,000 writes per second) and the relational database PostgreSQL to store structured data such as device status and maintenance records. Feature engineering unit 202 runs a convergent cross-mapping (CCM) algorithm, using COD as the target variable, to analyze its causal relationship with pH, ​​DO, and NH3-N: setting the embedding dimension E=5 and the time delay τ=3, the convergence coefficient (ρ>0.85 indicates a valid causal relationship) is calculated, verifying that the causal strength of NH3-N on COD is 0.92, significantly higher than the 0.67 of traditional Pearson correlation analysis.

[0035] 3. Bidirectional Neural Network Modeling The data prediction module 300 constructs a bidirectional neural network incorporating LSTM and BiGRU. The input layer receives water quality data from the past two hours (time step = 120), and the hidden layer has 128 neurons, outputting predicted COD and NH3-N values ​​for the next hour. Model training uses the Adam optimizer with an initial learning rate of 0.001 and a batch size of 64. The 30-day dataset is divided into a training set (80%) and a test set (20%). A dynamic weight adjustment mechanism is introduced during training, increasing the weight of the loss function for samples with prediction errors exceeding a threshold (e.g., COD > 15 mg / L) to improve the model's adaptability to abnormal operating conditions.

[0036] 4. Visualization and Comparison Testing The visualization module 400 integrates the Tableau engine to generate real-time data dashboards, predictive trend curves, and causal network topology diagrams. To quantify the system's advantages, two sets of comparative experiments were set up: Experimental group: Employing the entire workflow of this system (edge ​​computing + CCM feature selection + bidirectional neural network) Control group: Using a traditional SCADA system (raw data transmission + linear regression prediction) During the test, three typical scenarios were simulated: normal operating conditions, sudden change in influent load (COD increased by 40% instantaneously), and equipment failure (aerator power decreased by 30%), and key performance indicators were recorded.

[0037] II. Experimental Data Tables

[0038] III. Data Analysis and Conclusions 1. Sensor network performance optimization As shown in Table 1, the sensor network of this system significantly outperforms traditional equipment in terms of sampling frequency (60 times / minute vs. 10 times / minute) and anti-interference rate (98.7% vs. 82.4%). The edge computing unit (Table 2) reduces the amount of water quality parameter data by 65% ​​through adaptive filtering and compression algorithms, while ensuring a processing latency of less than 20 ms, thus solving the data packet loss problem caused by bandwidth limitations in traditional systems (completeness rate 99.5% vs. 91.2%).

[0039] 2. Advantages of causal feature extraction Table 3 shows that the CCM algorithm improves the effectiveness of causal relationship identification by 35-50%. For example, the causal strength between NH3-N and COD increases from the Pearson coefficient of 0.67 to 0.92. This provides a theoretical basis for the feature engineering unit 202 to screen key predictive variables and avoids the misselection of features caused by neglecting the nonlinear relationship of time series in traditional methods.

[0040] 3. Robustness verification of the prediction model In the three test scenarios in Table 4, the standardized root mean square error (NRMSE) of this system was lower than that of the control group. Especially in the load mutation scenario, the COD prediction error was only 9.1%, which is 66.6% lower than that of traditional linear regression. This is due to the joint modeling of historical / future information by the bidirectional neural network and the targeted learning of abnormal data by the dynamic weight adjustment mechanism.

[0041] 4. System-level performance improvement Regarding resource consumption (Table 5), this system reduces bandwidth requirements by 66% (18 vs 53 Mbps) and saves 60% of storage space through edge preprocessing and hybrid storage optimization. Operational response metrics (Table 6) further demonstrate that the causal topology graph based on the visualization module 400 improves fault location accuracy to 93%, a 38.8% improvement over traditional systems, while reducing anomaly detection latency to 2.1 minutes, meeting real-time requirements.

[0042] Conclusion: Through multi-dimensional data verification, this embodiment demonstrates that the system overcomes existing technical bottlenecks in terms of data acquisition density, feature interpretability, prediction accuracy, and operational efficiency. Particularly under dynamic operating conditions, its bidirectional neural network model error is controlled within 12%, more than three times higher than traditional methods, providing reliable technical support for the intelligent optimization of wastewater treatment processes.

[0043] In the description of this invention, it should be noted that the terms "center," "upper," "lower," "left," "right," "vertical," "horizontal," "inner," and "outer," etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are used only for the convenience of describing the invention and for simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on the invention. The terms "first," "second," and "third" are used for descriptive purposes only and should not be construed as indicating or implying relative importance. Furthermore, unless otherwise explicitly specified and limited, the terms "installed," "connected," and "linked" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal connection of two components. Those skilled in the art can understand the specific meaning of the above terms in this invention based on the specific circumstances.

[0044] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A remote wastewater treatment system based on a big data model, characterized in that, It includes an information collection module (100), a data processing module (200), a data prediction module (300), and a visualization module (400). The information collection module (100) includes a sensor network unit (101) and an edge signal computing unit (102). The sensor network unit (101) includes sensors deployed at the location of the sewage treatment facility. The edge signal computing unit (102) is used to preprocess the data collected by the sensor network unit (101). The data processing module (200) includes a hybrid storage unit (201) and a feature engineering unit (202). The hybrid storage unit (201) includes a time-series database and a relational database. The feature engineering unit (202) uses a convergent cross-mapping algorithm to verify the causal relationship between water quality parameters. The data prediction module (300) is used to establish a bidirectional neural network, collect data from the edge signal calculation unit (102) after preprocessing, and predict expected water quality indicators; The visualization module (400) collects the output data of the hybrid storage unit (201) and the data prediction module (300) and displays it in the form of images.

2. The remote sewage treatment system based on a big data model according to claim 1, characterized in that... The sensor network unit (101) includes sensors deployed at the location of the wastewater treatment facility, including: The sensor network unit (101) deploys sensors at the inlet, biological treatment tank, sedimentation tank and outlet of the wastewater treatment facility; A three-dimensional point layout is adopted in the biochemical reaction tank, with at least three detection layers set along the vertical direction, and composite sensors distributed in a honeycomb grid in each layer; An ultraviolet absorption spectroscopy detection component is installed at the water outlet.

3. The remote sewage treatment system based on a big data model according to claim 1, characterized in that, The edge signal calculation unit (102) is used to preprocess the data collected by the sensor network unit (101), including: Edge-cloud data collaborative control equation: ; Where ε(t) represents the cooperative effect, t is the synchronization timestamp, ∂Ω is the closed path formed by the sensor network in the complex plane, and F edge (z) is the feature function of the preprocessed data of the edge node z, G cloud Let Re(⋅) be the analytic extension function of the corresponding node in the cloud, where Re(⋅) takes the real part of the complex function, Li2(⋅) is the double logarithm function, and ρ sync Let ζ be the data synchronization rate between the edge and the cloud, ζ be the network jitter coefficient, Si(⋅) be the sine integral function, and δ be the data synchronization rate between the edge and the cloud. latency This represents the transmission delay in milliseconds.

4. The remote sewage treatment system based on a big data model according to claim 3, characterized in that, The data prediction module (300) is used to establish a bidirectional neural network, collect preprocessed data from the edge signal calculation unit (102), and predict expected water quality indicators, including: Water quality prediction-control coupling equations: ; Where P(t) represents the system control sensitivity, Proj g (⋅) is the projection operator on the Lie algebra g, ∇ θ L BiLSTM Let J be the gradient of the loss function of the bidirectional neural network, so(3) be the three-dimensional rotation group norm, and J be the gradient of the loss function. CCM T Let Det(⋅) be the convergent cross-mapping matrix, and Σ be the determinant. sensor For the sensor data covariance matrix, ent KL ν is the divergence entropy between the predicted and measured values, and ν is the equipment aging correction factor.

5. A remote wastewater treatment system based on a big data model according to claim 1, characterized in that, The hybrid storage unit (201) includes: Time-series databases, relational databases, and distributed file systems are used for hierarchical storage of real-time water quality data, equipment metadata, and unstructured data.

6. The remote sewage treatment system based on a big data model according to claim 1, characterized in that, The visualization module (400) collects the output data from the hybrid storage unit (201) and the data prediction module (300) and displays it in image form, including: The output parameters of the visualization module (400) are sent to the PLC controller, and an alarm threshold is set. When the wastewater quality value exceeds the wastewater quality value predicted by the data prediction module (300), an email notification is triggered.

7. A remote sewage treatment system based on a big data model according to claim 5, characterized in that, The time-series database, relational database, and distributed file system include: Use a time-series database to store real-time streaming data from sensors; Relational databases store device metadata and process parameters; The distributed file system stores video surveillance and equipment maintenance log data.

8. A remote wastewater treatment system based on a big data model according to claim 1, characterized in that, The feature engineering unit (202) uses a convergent cross-mapping algorithm to verify the causal relationship between water quality parameters, including: The feature engineering unit (202) calculates the spatial gradient of adjacent sensor data, extracts time-series statistical features through a sliding window, and uses a convergent cross-mapping algorithm to verify the causal relationship between water quality parameters. The time-series statistical features include mean, variance, trend slope, and permutation entropy.

9. An electronic device, characterized in that... include: One or more processors; A storage device on which one or more programs are stored; When the one or more programs are executed by the one or more processors, the one or more processors implement the system as described in any one of claims 1-8.

10. A computer-readable storage medium having executable instructions stored thereon, characterized in that... When executed by the processor, this instruction causes the processor to implement the system as described in any one of claims 1-8.

Images