Medical sewage intelligent data terminal based on edge computing

Abstract

The present application relates to sewage monitoring technical field, specifically, it is a kind of medical sewage intelligent data terminal based on edge computing.It includes multi-source data fusion unit, edge computing unit and overproof early warning unit, multi-source data fusion unit forms water quality perception network, makes edge computing unit based on water quality perception network constructs the digital image of sewage state, double-index auxiliary verification outputs sewage state, and makes overproof early warning unit according to matching result, generates multi-point early warning signal, the present application quantitatively quantifies the difference of upper and lower layer pollutants, generates the quantitative evaluation index of layered or mixed sewage, matches different processing strategies, is beneficial to the sewage of different stratification for separate treatment, improves sewage treatment effect, and when output uniform dissolution state, then the output data of fusion multiple sewage monitors determines water quality parameter mean, and the mean of the monitoring data of multiple different arrangement points is verified sewage index more accurately.

Description

Technical Field

[0001] This invention relates to the field of wastewater monitoring technology, and more specifically, to a smart data terminal for medical wastewater based on edge computing. Background Technology

[0002] The "Medical Wastewater Intelligent Data Terminal" is an IoT-based intelligent monitoring and data management device applied to medical wastewater treatment scenarios. Currently, wastewater (including drug residues and cleaning solutions) generated during medical use is collected in collection tanks. The terminal then monitors key indicators of medical wastewater treatment in real-time, online, and automatically, transmitting the data to the cloud or management platform. This enables intelligent monitoring and management of the wastewater treatment process, allowing managers to monitor whether wastewater treatment meets standards and whether equipment is operating normally.

[0003] However, current intelligent data terminals for medical wastewater mostly rely on a single data point to determine whether wastewater exceeds standards, resulting in limited accuracy. Furthermore, wastewater exists in multiple states, including static stratified states (with impurities floating on top and settling at the bottom) and uniformly dissolved states. Currently, the monitoring, discharge, and treatment processes do not differentiate between these states, leading to poor targeting of intelligent wastewater treatment. Therefore, we propose an intelligent data terminal for medical wastewater based on edge computing. Summary of the Invention

[0004] The purpose of this invention is to provide a smart data terminal for medical wastewater based on edge computing, so as to solve the following problems mentioned in the background art:

[0005] Firstly, there is a challenge in real-time monitoring of medical wastewater: traditional monitoring methods rely on a single sensor or manual sampling, which makes it difficult to capture dynamic changes in wastewater in real time (such as stratification or uneven distribution of pollutants), leading to delays or misjudgments in wastewater treatment. Furthermore, there is a lack of quantitative identification of static stratification phenomena in wastewater (such as stratification of sedimented or floating pollutants), affecting subsequent accurate treatment decisions.

[0006] Secondly, the limitations of data fusion and analysis: multi-sensor data (concentration, COD, pH, etc.) are processed independently, without effectively integrating spatiotemporal correlations, making it difficult to construct a global profile of the wastewater status.

[0007] Third, the precision of wastewater treatment is insufficient: the discharge strategies for uniformly mixed and stratified wastewater are not differentiated, resulting in excessive dosing (increasing costs) or incomplete treatment (environmental risks).

[0008] To achieve the above objectives, the present invention provides a medical wastewater intelligent data terminal based on edge computing, including a multi-source data fusion unit, an edge computing unit, and an exceedance early warning unit. It dynamically identifies the uniform dissolution or static stratification state of wastewater, quantifies the differences between upper and lower layers of pollutants, generates quantitative evaluation indicators for stratified or mixed wastewater, and matches differentiated treatment strategies.

[0009] The multi-source data fusion unit is used to assemble multiple wastewater monitors at preset locations to form a spatially distributed water quality sensing network, including inlet locations.

[0010] The edge computing unit is used to extract the output data of multiple sewage monitors to construct a digital profile of the sewage status and output the sewage status. The sewage status includes static stratification state and uniform dissolution state. If the static stratification state is output, the water quality parameters of multiple deployment points are output and matched with the discharge index. If the uniform dissolution state is output, the output data of multiple sewage monitors are merged to determine the average water quality parameter and match the average water quality parameter with the discharge index.

[0011] The above-mentioned early warning unit is used to receive signals that the water quality parameters and emission indicators do not match at multiple locations, and signals that the average water quality parameters do not match the emission indicators, and to issue early warning notifications.

[0012] As a further improvement to this technical solution, the multi-source data fusion unit includes wastewater monitor A, wastewater monitor B, and wastewater monitor C. Wastewater monitor A, wastewater monitor B, and wastewater monitor C form a water quality sensing network, and each is composed of a concentration sensor and a water quality sensor. The concentration sensor is used to obtain the wastewater concentration value, and the water quality sensor is used to obtain water quality parameters, including COD, SS, pH, and conductivity.

[0013] As a further improvement to this technical solution, the multi-source data fusion unit also includes a stagnation monitoring box. The side wall of the stagnation monitoring box near the top of the inner cavity is connected to a water inlet. The side wall of the stagnation monitoring box is connected to a bottom drainage end and an upper drainage end. The bottom drainage end and the upper drainage end are respectively located in the upper and lower regions divided at the center line of the inner cavity of the stagnation monitoring box.

[0014] Wastewater monitor A is arranged at the inlet end of the stagnation monitoring box as the inlet point, and wastewater monitors B and C are arranged near the bottom drainage end and the top drainage end, respectively, so that wastewater monitors B and C can respectively cover the monitoring of the wastewater status of the upper and lower areas.

[0015] As a further improvement to this technical solution, the edge computing unit includes a digital profile preset module and a sewage status output module;

[0016] The digital profiling preset module is used to preset wastewater state rules. When the direct concentration difference does not persist through time series stability detection, it outputs the following state rules for wastewater:

[0017] If a≈b≈c, the output wastewater will always be in a uniformly dissolved state without stratification.

[0018] If a > b > c or a < b < c, output the static stratification state of pollutants' sedimentation and floating trends;

[0019] Among them, the concentration value of wastewater monitor A is a, the concentration value of wastewater monitor B is b, and the concentration value of wastewater monitor C is c.

[0020] As a further improvement to this technical solution, the digital profile preset module also includes a concentration threshold matching module, which is used to set a concentration threshold based on the pollutant type, including the following:

[0021] Attitude 1: When |bc| > concentration threshold, output static stratification state. If b > c, the sediment in the lower layer is higher than the floating matter in the upper layer. If c > b, the floating matter in the upper layer is higher than the sediment in the lower layer, and there are still residual pollutants in the upper layer.

[0022] Posture 2: When |bc|≤concentration threshold, output a uniformly dissolved state.

[0023] As a further improvement to this technical solution, the wastewater status output module is used to train a neural network model based on historical data, construct a linear relationship between the trend of water quality parameter changes and the wastewater status, and combine the concentration values ​​and water quality parameter change trends of multiple wastewater monitors at multiple locations in the multi-source data fusion unit. Multiple concentration values ​​are input into the wastewater status rules to output the initial wastewater status, and the water quality parameter change trends are input into the neural network model to output the secondary wastewater status. If the initial wastewater status and the secondary wastewater status are consistent, the wastewater status is output; if they are inconsistent, a source tracing warning is issued.

[0024] As a further improvement to this technical solution, in the static layered state, the edge computing unit includes a dynamic weighting module and a layered quantization module;

[0025] The dynamic weighting module is used to standardize each water quality parameter and concentration value to form a single-factor pollution index, and to assign different weights to different water quality parameters and concentration values ​​according to the degree of pollutant hazard, the strictness of emission standards, and treatment cost factors, and to calculate the comprehensive pollution index.

[0026] The hierarchical quantification module is used to set emission indicators. The comprehensive pollution index of wastewater monitor B and wastewater monitor C is calculated sequentially through the dynamic weight module. The comprehensive pollution index is compared with the emission indicators. If they match the emission indicators, an emission signal is issued. If they do not match the emission indicators, a processing signal is issued.

[0027] As a further improvement to this technical solution, in the uniformly dissolved state, the edge computing unit includes a mean calculation module. The mean calculation module is used to calculate the comprehensive pollution index of wastewater monitors A, B, and C respectively, based on the weighted comprehensive pollution index values ​​output by the water quality parameter quantification module. The mean of the three comprehensive pollution indices is calculated, and the mean is compared with the emission index. If the emission index matches, an emission signal is issued; otherwise, a processing signal is issued.

[0028] As a further improvement to this technical solution, the over-standard early warning unit includes a display screen showing the status of wastewater, a green indicator light for receiving discharge signals, and a red indicator light for receiving processing signals.

[0029] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0030] In this intelligent data terminal for medical wastewater based on edge computing, the wastewater status is determined by the edge computing unit. When the static stratification status is output, the water quality parameters and discharge indicators of multiple deployment points are output respectively, which is conducive to the individual treatment of wastewater of different stratifications and improves the wastewater treatment effect.

[0031] When the output is in a uniformly dissolved state, the output data of multiple wastewater monitors are combined to determine the average value of water quality parameters. The wastewater index is more accurate by verifying the average value of monitoring data from multiple monitoring points with different locations. Attached Figure Description

[0032] Figure 1 This is a schematic diagram of the overall structure of the present invention;

[0033] Figure 2 This is a schematic diagram of the stagnation monitoring box of the present invention.

[0034] The meanings of the labels in the diagram are as follows:

[0035] 1. Multi-source data fusion unit; 10. Stagnation monitoring box; 11. Water inlet; 12. Bottom drainage end; 13. Upper drainage end;

[0036] 2. Edge computing unit; 3. Over-limit early warning unit. Detailed Implementation

[0037] 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.

[0038] Example 1: Please refer to Figures 1-2 As shown, this embodiment provides a smart data terminal for medical wastewater based on edge computing, including a multi-source data fusion unit 1, an edge computing unit 2, and an over-standard early warning unit 3;

[0039] Step 1: The multi-source data fusion unit 1 is used to pre-arrange multiple wastewater monitors at designated locations to form a spatially distributed water quality sensing network. The locations include inlet points (for monitoring the quality of medical raw water entering the stagnation monitoring box 10) and stratification line points (for detecting whether the wastewater is in a stratified state and the differences in water quality between the upper and lower layers). The data collected by the multiple wastewater monitors working together has timestamps, location tags, concentration values, and parameter types, which facilitates subsequent analysis.

[0040] Wherein: the multi-source data fusion unit 1 includes wastewater monitor A, wastewater monitor B and wastewater monitor C. Wastewater monitor A, wastewater monitor B and wastewater monitor C form a water quality sensing network, and each is composed of a concentration sensor and a water quality sensor. The concentration sensor is used to obtain the wastewater concentration value, and the water quality sensor is used to obtain water quality parameters, including COD, SS (suspended solids), pH and conductivity.

[0041] Specifically, the concentration sensor uses an optical sensor. Wastewater flows through the sensor's detection window, and an ultraviolet LED emits 254nm light waves that penetrate the wastewater. Organic matter absorbs the ultraviolet light, and a photodetector measures the attenuation of the transmitted light intensity. According to Beer-Lambert's law, the absorbance is converted into the COD value (e.g., absorbance 0.5 → COD = 200mg / L). Note: The optical sensor needs to be cleaned regularly to prevent interference from dirt, which is known to those skilled in the art and will not be elaborated here. Therefore, wastewater monitors A, B, and C can collect concentration values ​​near multiple locations. Wastewater monitor A obtains concentration value a, wastewater monitor B obtains concentration value b, and wastewater monitor C obtains concentration value c.

[0042] Water quality sensors include:

[0043] The pH sensor uses a glass electrode method, where the pH glass electrode is in contact with the wastewater. A reference electrode provides a stable potential reference, and H⁺ ions pass through the sensitive membrane to generate a potential difference (mV), which is converted into a pH value (e.g., pH=6.5) using the Nernst equation. A temperature sensor is also included to correct the pH value. Temperature changes are sensed through a platinum resistance thermometer (RTD) or a thermocouple, and a resistance / voltage signal (e.g., 25℃) is output.

[0044] Dissolved oxygen (DO) sensors utilize the Clarke electrode method: oxygen molecules permeate through the membrane and undergo a reduction reaction with the cathode (O2 + 4e⁻ → 2O²⁻), and the current intensity is directly proportional to the DO concentration (e.g., DO = 5 mg / L).

[0045] The conductivity sensor applies alternating current between two electrodes and measures the resistance value (inversely proportional to conductivity) to reflect the ion content of wastewater (e.g., conductivity = 1500 μS / cm).

[0046] Specifically, such as Figure 2 As shown, the multi-source data fusion unit 1 also includes a stagnation monitoring box 10. The side wall of the stagnation monitoring box 10 near the top of the inner cavity is connected to a water inlet 11. The side wall of the stagnation monitoring box 10 is connected to a bottom drainage end 12 and an upper drainage end 13. The bottom drainage end 12 and the upper drainage end 13 are respectively located in the upper and lower regions divided at the center line of the inner cavity of the stagnation monitoring box 10.

[0047] Wastewater monitor A is arranged at the inlet end 11 of the stagnation monitoring box 10 as the inlet point, and wastewater monitor B and wastewater monitor C are arranged near the bottom drainage end 12 and the upper drainage end 13 respectively, so that wastewater monitor B and wastewater monitor C can respectively cover the monitoring of the wastewater status of the upper and lower areas.

[0048] Therefore, medical wastewater is discharged from the inlet 11 into the stagnation monitoring box 10. During the discharge process, the wastewater monitor A monitors the water quality parameters in real time (at this time, the water is kept flowing as it enters from the inlet 11, so that the wastewater is in a relatively uniform mixed state during monitoring). When the stagnation monitoring box 10 is full of wastewater (a level gauge can be installed to determine the amount of wastewater entering, so as to determine whether the wastewater is full), it is allowed to stand still for a certain period of time to allow the wastewater to stratify, which is convenient for subsequent digital profiles or water quality parameters to be formed based on the output data of each deployment point.

[0049] The static time can be obtained through training based on past data, such as 2-4 hours.

[0050] Step 2: The edge computing unit 2 is used to construct a digital profile of the sewage state based on the water quality sensing network. The sewage state is verified by dual indicators. The sewage state includes static stratification state and uniform dissolution state. If the static stratification state is output, the comprehensive pollution index of multiple deployment points is output in sequence and matched with the emission index in sequence. If the uniform dissolution state is output, the average comprehensive pollution index of multiple sewage monitors is merged and matched with the average comprehensive pollution index and the emission index.

[0051] One of its purposes is to determine whether the sewage in the stagnation monitoring box 10 has stratified after a certain period of time, so as to facilitate subsequent sewage treatment and discharge based on the sewage status. The edge computing unit 2 includes a digital profile preset module and a sewage status output module.

[0052] The digital profiling preset module is used to preset wastewater state rules. It outputs the following state of the wastewater: when the direct concentration difference is not continuous (if the concentration difference between upper and lower layers continues for 2 hours (e.g., |bc|>Δ for 5 consecutive samplings), it is determined to be stratified, avoiding misjudgment due to instantaneous interference:

[0053] If a≈b≈c, the output wastewater will always be in a uniformly dissolved state without stratification.

[0054] If a > b > c or a < b < c, output the static stratification state of pollutants' sedimentation and floating trends;

[0055] Among them, the concentration value of wastewater monitor A is a, the concentration value of wastewater monitor B is b, and the concentration value of wastewater monitor C is c.

[0056] Where a1≈b1≈c1 indicates that the concentration value output by wastewater monitor A is approximately equal to the concentration value output by wastewater monitor B, and approximately equal to the concentration value output by wastewater monitor C. "Approximately equal" indicates that it is within the concentration threshold range. Therefore, to improve the accuracy of wastewater status judgment, the digital profile preset module also includes a concentration threshold matching module. The concentration threshold matching module is used to dynamically set the concentration threshold based on departmental information, including the following:

[0057] Attitude 1: When |bc| > concentration threshold, output static stratification state. If b > c, the sediment in the lower layer is higher than the floating matter in the upper layer. If c > b, the floating matter in the upper layer is higher than the sediment in the lower layer, and there are still residual pollutants in the upper layer.

[0058] Posture 2: When |bc|≤concentration threshold, output uniform dissolution state to further verify the state of wastewater, indicating that all pollutants are stably dissolved, or that the wastewater has undergone thorough mixing treatment (such as chemical coagulation).

[0059] It is worth noting that in medical wastewater treatment scenarios, the types, concentrations, and risk levels of pollutants generated by different departments vary significantly. To determine the concentration threshold, it is necessary to combine the department's business characteristics, pollutant emission patterns, and regulatory requirements to achieve accurate matching. Specifically, three months of departmental emission data should be collected, and cluster analysis (such as K-means) should be used to verify the rationality of the threshold. A department-pollutant mapping table should be established (where, when wastewater from multiple departments is mixed, such as when sharing pipelines, a weighted threshold should be used).

[0060] Predefine the initial concentration threshold for each type of department, install an RFID tag on each sewage collection tank to indicate the department type, and automatically call the corresponding threshold template after the terminal reads the RFID tag;

[0061] The concentration threshold needs to be dynamically adjusted based on actual operating data, with the following adjustment factors added:

[0062] Flow correction: If the wastewater flow rate of a department is <5L / min (such as dental), the SS threshold will be reduced by 20% (Δ[SS]=15→12mg / L).

[0063] Time adjustment: The laboratory will temporarily increase the COD threshold by 30% during the 9:00-11:00 (centralized testing period) (Δ[COD]=20→26mg / L).

[0064] Then, the wastewater status output module is used to train a neural network model based on historical data, construct a linear relationship between the water quality parameter change trend and the wastewater status, and combine the concentration values ​​and water quality parameter change trends of multiple wastewater monitors at multiple locations in the multi-source data fusion unit 1. Multiple concentration values ​​are input into the wastewater status rules to output the initial wastewater status. The water quality parameter change trend is input into the neural network model to output the secondary wastewater status. If the initial wastewater status and the secondary wastewater status are consistent, the wastewater status is output. If they are inconsistent, a source tracing warning is issued. Specifically, if the concentration value output by wastewater monitor A is a1, the concentration value output by wastewater monitor B is b1, and the concentration value output by wastewater monitor C is c1, if a1≈b1≈c1, the wastewater is output as a uniformly dissolved state without stratification. If a1>b1>c1 or a1<b1<c1, the pollutants are output as a static stratified state with sedimentation and floating trends, which facilitates the judgment of stratification status based on the concentration value.

[0065] Furthermore, when training a neural network model based on historical data, features from the historical data are input, including:

[0066] Time-series water quality parameters: sliding window data of parameters such as COD, SS, pH, DO, and conductivity over the past 1 hour (sampling frequency 5 minutes, 12 time points in total).

[0067] Characteristics of changing trends: mean, variance, slope (linear fitting), range (max-min).

[0068] Environmental auxiliary data: water temperature, flow rate (affecting pollutant diffusion);

[0069] Wastewater state categories are categorized by historical manual testing or high-precision instruments, including: uniform mixing, sedimentation-type stratification, and flotation-type stratification;

[0070] A lightweight 1D-CNN+LSTM hybrid network is selected, suitable for time-series data classification, to form a neural network model (training process, including loss function: categorical cross-entropy; optimizer: Adam (learning rate = 0.001). Data augmentation: Gaussian noise is added (simulating sensor error) to improve robustness. Validation method: 20% test set is reserved, accuracy must be >90% (this technique is known to those skilled in the art and will not be described in detail here), so that the water quality parameter change trend output from the multi-source data fusion unit 1 is input into the neural network model to output the wastewater status;

[0071] It facilitates dual verification of wastewater status, improving accuracy. In case of inconsistency, after source tracing and early warning, conflict cause analysis tools can be used: if the rule judges it as stratification but the model judges it as mixing → check if there has been a recent change in flow rate (a brief stirring caused a mixing illusion); if the model judges it as stratification but the rule judges it as mixing → analyze whether new pollutants have been added (such as grease not included in the rule base).

[0072] The second purpose is to process the static stratification state and the uniform dissolution state separately. In the static stratification state, the edge computing unit 2 includes a dynamic weighting module and a stratification quantization module.

[0073] The dynamic weighting module is used to standardize each water quality parameter and concentration value to form a single-factor pollution index. Based on the degree of hazard of pollutants, the strictness of emission standards, and treatment cost factors, different weights are assigned to different water quality parameters and concentration values ​​to calculate the comprehensive pollution index. For example, COD: 0.2, SS: 0.15, NH3-N: 0.2, pH: 0.1, DO: 0.1, conductivity: 0.1, TP: 0.05, TN: 0.1, and the comprehensive pollution index ∈ [0,1]. The closer it is to 1, the more serious the pollution.

[0074] Specifically, based on the degree of hazard of pollutants, the stringency of emission standards, and treatment cost factors, different weights are assigned to different water quality parameters and concentration values ​​using a three-dimensional weighting model. This model includes a weight for the degree of hazard (based on pollutant toxicology data classification), a weight for the stringency of emission standards (the higher the value, the more stringent the standard), and a weight for treatment costs (based on the normalization of departmental pollutant treatment costs). In this process, a basic pollutant database is first constructed, and the three-dimensional weighting calculation rules are preset (based on historical data fitting and training of a neural network model), and then the weights are calculated comprehensively.

[0075] The hierarchical quantification module is used to set emission indicators (such as a comprehensive pollution index ≤ 0.7 indicating compliance). The comprehensive pollution index of wastewater monitor B and wastewater monitor C is calculated sequentially through the dynamic weight module. The comprehensive pollution index is compared with the emission indicators. If they match the emission indicators, an emission signal is issued. If they do not match the emission indicators, a processing signal is issued.

[0076] When pollutants meet emission limits, a discharge signal is issued, allowing wastewater to be discharged directly without treatment. During discharge, wastewater from the upper layer is first discharged through the upper drainage end 13, and then wastewater from the lower layer is discharged through the bottom drainage end 12. Conversely, when a treatment signal is issued, wastewater from the upper layer is discharged through the corresponding upper drainage end 13, starting from the bottom of the upper layer and proceeding outwards. Wastewater from the upper layer is treated separately. After the upper layer discharge is completed, wastewater from the lower layer is discharged through the bottom drainage end 12 for separate treatment. This allows for separate treatment of wastewater from different layers, improving wastewater treatment efficiency.

[0077] In a uniformly dissolved state, the edge computing unit 2 includes a mean calculation module. The mean calculation module is used to calculate the comprehensive pollution index of wastewater monitors A, B, and C respectively, based on the weighted comprehensive pollution index values ​​output by the water quality parameter quantification module. The mean of the three comprehensive pollution indices is calculated and compared with the discharge index. If the discharge index matches, a discharge signal is issued, and all wastewater from the wastewater stagnation monitoring box 10 can be directly discharged through the bottom drainage end 12. If the discharge index does not match, a treatment signal is issued, and all wastewater from the wastewater stagnation monitoring box 10 can also be treated by discharging it through the bottom drainage end 12. Verifying the wastewater index using the mean method is more accurate.

[0078] Step 3: The over-standard early warning unit 3 is used to receive the matching results of the edge computing unit 2 and generate multi-point early warning signals.

[0079] Specifically, the multi-point warning signal of the above-standard warning unit 3 includes a display screen showing the status of sewage, a green indicator light for receiving discharge signals, and a red indicator light for receiving processing signals. This allows users to intuitively judge the status of sewage through the display screen and thus determine different sewage discharge methods. The green and red indicator lights help users intuitively determine whether to choose direct discharge or sewage discharge for treatment.

[0080] The information on the display screen is displayed in layers.

[0081] First line: Real-time status (uniform mixing / sedimentation / floating layering)

[0082] Second row: Simplified table of WPI index and water quality parameters (e.g., COD: 85 | SS: 120 | pH: 6.5)

[0083] Third line: Treatment recommendations (direct discharge / requires flotation treatment / emergency shutdown);

[0084] Through a multi-level alarm linkage mechanism

[0085] 1. Level 1 warning (0.7) <WPI≤1.0)

[0086] Indicator light: Red slow flashing (1Hz);

[0087] Screen notification: Yellow highlighted "Slightly exceeded standard, manual confirmation required";

[0088] Action: Provides a prompt only, without interrupting the process.

[0089] 2. Level 2 Warning (1.0) <WPI≤1.5);

[0090] Indicator light: Red flashing (2Hz) + buzzer sound;

[0091] The screen displays a flashing red message: "Moderate exceedance! Pending processing."

[0092] 3. Level 3 warning (WPI>1.5)

[0093] Indicator light: Solid red + continuous buzzer;

[0094] The screen displayed a message in red text with white text: "Severe pollution! System has been shut down!"

[0095] Action: Cut off the power supply and report to the environmental protection platform.

[0096] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely preferred examples and are not intended to limit the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the present invention as claimed. The scope of protection of the present invention is defined by the appended claims and their equivalents.

Claims

1. A smart data terminal for medical wastewater based on edge computing, characterized in that: It includes a multi-source data fusion unit (1), an edge computing unit (2), and an over-standard early warning unit (3); The multi-source data fusion unit (1) is used to pre-arrange multiple sewage monitors at designated locations to form a spatially distributed water quality sensing network. The water quality sensing network outputs sewage concentration values ​​and water quality parameters. The edge computing unit (2) is used to construct a digital profile of the sewage state based on the water quality sensing network. The sewage state is verified by dual indicators. The sewage state includes static stratification state and uniform dissolution state. If the static stratification state is output, the comprehensive pollution index of multiple deployment points is output in sequence and the emission index is matched in sequence. If the uniform dissolution state is output, the average comprehensive pollution index of multiple sewage monitors is integrated and the average comprehensive pollution index is matched with the emission index. The over-standard early warning unit (3) is used to receive the matching result of the edge computing unit (2) and generate a multi-point early warning signal; The edge computing unit (2) includes a wastewater status output module, which is used to train a neural network model based on historical data, construct a linear relationship between the water quality parameter change trend and the wastewater status, and combine the concentration values ​​and water quality parameter change trends of multiple wastewater monitors at multiple locations in the multi-source data fusion unit (1) to input multiple concentration values ​​into the wastewater status rules and output the initial wastewater status. The water quality parameter change trend is input into the neural network model and output the secondary wastewater status. If the initial wastewater status and the secondary wastewater status are consistent, the wastewater status is output. If they are inconsistent, a source tracing warning is issued.

2. The medical wastewater intelligent data terminal based on edge computing according to claim 1, characterized in that: The multi-source data fusion unit (1) includes wastewater monitor A, wastewater monitor B and wastewater monitor C. Wastewater monitor A, wastewater monitor B and wastewater monitor C form a water quality sensing network and are all composed of concentration sensors and water quality sensors. The concentration sensors are used to obtain wastewater concentration values ​​and the water quality sensors are used to obtain water quality parameters, including COD, SS, pH and conductivity.

3. The medical wastewater intelligent data terminal based on edge computing according to claim 2, characterized in that: The multi-source data fusion unit (1) also includes a stagnation monitoring box (10). The stagnation monitoring box (10) has a water inlet (11) connected to the side wall near the top of the inner cavity. The side wall of the stagnation monitoring box (10) has a bottom drainage end (12) and an upper drainage end (13) connected to it. The bottom drainage end (12) and the upper drainage end (13) are respectively located in the upper and lower regions divided at the center line of the inner cavity of the stagnation monitoring box (10). Wastewater monitor A is arranged at the inlet end (11) of the stagnation monitoring box (10) as the inlet point, and wastewater monitor B and wastewater monitor C are arranged near the bottom drainage end (12) and the upper drainage end (13) respectively, so that wastewater monitor B and wastewater monitor C can respectively cover the monitoring of the wastewater status of the upper and lower areas.

4. The medical wastewater intelligent data terminal based on edge computing according to claim 3, characterized in that: The edge computing unit (2) includes a digital profile preset module; The digital profile preset module is used to preset wastewater state rules. When the concentration difference does not persist through time series stability detection, the wastewater state rules are output in the following manner: If a≈b≈c, the output wastewater will always be in a uniformly dissolved state without stratification. If a > b > c or a < b < c, output the static stratification state of pollutants' sedimentation and floating trends; Among them, the concentration value of wastewater monitor A is a, the concentration value of wastewater monitor B is b, and the concentration value of wastewater monitor C is c.

5. The medical wastewater intelligent data terminal based on edge computing according to claim 4, characterized in that: The digital profile preset module also includes a concentration threshold matching module, which is used to dynamically set concentration thresholds based on departments, including the following: Attitude 1: When |bc| > concentration threshold, output static stratification state. If b > c, the sediment in the lower layer is higher than the floating matter in the upper layer. If c > b, the floating matter in the upper layer is higher than the sediment in the lower layer, and there are still residual pollutants in the upper layer. Posture 2: When |bc|≤concentration threshold, output a uniformly dissolved state.

6. The medical wastewater intelligent data terminal based on edge computing according to claim 5, characterized in that: In the static hierarchical state, the edge computing unit (2) includes a dynamic weighting module and a hierarchical quantization module; The dynamic weighting module is used to standardize each water quality parameter and concentration value to form a single-factor pollution index, and to establish a three-dimensional weighting model to assign different weights to different water quality parameters and concentration values ​​based on the degree of pollutant hazard, the strictness of emission standards, and treatment cost factors, and to calculate the comprehensive pollution index. The hierarchical quantification module is used to set emission indicators. The comprehensive pollution index of wastewater monitor B and wastewater monitor C is calculated sequentially through the dynamic weight module. The comprehensive pollution index is compared with the emission indicators. If they match the emission indicators, an emission signal is issued. If they do not match the emission indicators, a processing signal is issued.

7. The medical wastewater intelligent data terminal based on edge computing according to claim 6, characterized in that: In a uniformly dissolved state, the edge computing unit (2) includes a mean calculation module. The mean calculation module is used to calculate the comprehensive pollution index of wastewater monitor A, wastewater monitor B and wastewater monitor C respectively output by the water quality parameter quantification module, calculate the mean of the three comprehensive pollution indices, and compare the mean with the emission index. If the emission index matches, an emission signal is issued; if the emission index does not match, a processing signal is issued.

8. The intelligent medical wastewater data terminal based on edge computing according to claim 1, characterized in that: The multi-point warning signals of the above-standard warning unit (3) include a display screen showing the status of sewage, a green indicator light for receiving discharge signals, and a red indicator light for receiving processing signals.

Images