Floating island type microbial fuel cell sensor for on-line monitoring of water pollution event

Abstract

The invention discloses a floating island type microbial fuel cell sensor for monitoring a water pollution event on line, and belongs to the technical field of environmental protection. The sensor comprises a sinking bucket, a floating island, a support frame, an anode, a cathode, an external resistor, a data collector, a solar panel and a monitoring terminal. According to the sensor, the anode is buried in the soil in the sinking bucket, and the sinking bucket is suspended below the floating island, so that the sensor does not need to submerge into the bottom of the wetland to embed the anode, and is suitable for monitoring the deep wetland; the device is provided with a data collector and a solar panel, so that electric signal data can be transmitted to a computer and a mobile phone terminal online in real time; by introducing a machine learning algorithm, a nonlinear mapping model between the output voltage signal of the microbial fuel cell and the pH value of the water body is established, so that the monitoring terminal can intelligently analyze the voltage data acquired in real time, thereby realizing high-precision, continuous and online monitoring of the pH value of the water body.

Description

TECHNICAL FIELD

[0001] The present application belongs to the field of environmental protection, and particularly relates to a floating island type microbial fuel cell sensor for online monitoring of water pollution events. BACKGROUND

[0002] Industrial production often discharges acidic and alkaline wastewater. At present, the monitoring of water pH mainly includes manual sampling analysis and online monitoring. Manual sampling analysis needs a sampling time interval of several days or even longer, which is difficult to find pollution in real time, and the labor and transportation costs are high; while online monitoring mainly uses glass electrodes which are easy to break, and the maintenance cost is high.

[0003] Patent CN120064417A discloses a method for monitoring acid-alkaline wastewater by using a sediment microbial fuel cell sensor matrix. The principle is that the ubiquitous electrogenic bacteria in wetland sediment decompose organic matter to generate electrons, which are transferred to the anode of the sediment microbial fuel cell, and the electrons flow from the anode in the sediment to the cathode in the overlying water through the wire, and react with H + and O2 in water to generate H2O, thereby generating a voltage on the circuit. The voltage change amplitude caused by acid-alkaline wastewater is converted into the water pH value by algorithm, so that the sediment microbial fuel cell can be used as a sensor to monitor the discharge of acid-alkaline wastewater in real time. Further research by the inventor shows that adding an acidic solution to the water near the cathode can promote the cathode reaction, causing the voltage to rise rapidly; adding an alkaline solution can inhibit the cathode reaction, causing the voltage to drop rapidly, and the amplitude of voltage rise or drop is linearly related to pH, which means that the stronger the acidity, the greater the amplitude of voltage rise; the stronger the alkalinity, the greater the amplitude of voltage drop (Electroanalysis, 2025, 37:e70010).

[0004] On the one hand, this method is only applicable to wetland environments with shallow water depth, because the anode needs to be inserted into the wetland sediment, and when the water depth is deep, it will be very difficult to insert the anode into the sediment, and it is dangerous for personnel to dive. In addition, this method relies on the functional relationship between voltage change and water pH value, and converts the electrical signal into water pH value by algorithm, so it is difficult to adapt to the dynamic changes of temperature, ion concentration and biological activity in natural wetland environment, resulting in a decrease in detection accuracy. With the help of machine learning, a water pH recognition neural network model based on microbial fuel cell voltage data can be constructed, and through continuous training and adaptive updating, the nonlinear mapping relationship between voltage signal and pH value can be automatically learned, thereby improving the accuracy and environmental adaptability of water pH monitoring. SUMMARY

[0005] One objective of this invention is to provide a floating island-type microbial fuel cell sensor for online monitoring of water pollution events, comprising a submerged tank, a floating island, a support frame, an anode, a cathode, an external resistor, a data acquisition unit, a solar panel, and a monitoring terminal; The submersible tank is positioned below the floating island and submerged in the water; the floating island floats on the water surface; the support frame is positioned on the surface of the floating island, the external resistor and data acquisition device are fixed on the support frame, and the solar panel is positioned on top of the support frame. The submerged tank contains soil, and the anode is buried in the soil; the cathode is fixed to the bottom of the floating island and submerged in the water; the anode and cathode assembly are connected in series with an external resistor via wires to form a microbial fuel cell; the external resistor is connected in parallel with a data acquisition device, and the data acquisition device transmits the voltage data across the external resistor of the microbial fuel cell to a monitoring terminal.

[0006] Furthermore, the anode is made of stainless steel, graphite, or conductive carbon fiber (such as carbon felt, carbon cloth, or carbon brush).

[0007] Furthermore, the cathode is made of stainless steel or titanium and is in the shape of a mesh or sheet.

[0008] Furthermore, the material of the conductor is titanium.

[0009] Furthermore, the floating island has a through-hole in the middle for embedding a submersible bucket, making it an integrated device.

[0010] Furthermore, aquatic plants are planted in the soil of the submerged tank. These aquatic plants provide a continuous carbon source for the soil-generating bacteria, which not only helps to increase the voltage of the microbial fuel cell and maintain a stable voltage output over a long period, preventing the limited soil organic matter in the submerged tank from being depleted and unable to maintain power generation, but also beautifies the sensor's appearance and creates a landscape.

[0011] The method for online monitoring of water pollution events using the aforementioned sensors includes the following steps: Step 1: Float the floating island on the water surface, immerse the cathode below the water surface, and immerse the submerged bucket suspended below the floating island in the water. The submerged bucket is filled with soil. Bury the anode in the soil of the submerged bucket and fix the external resistor, data acquisition device and solar panel above the floating island. Step 2: Connect the anode and cathode in series with an external resistor via wires to form a microbial fuel cell. Connect the data acquisition unit in parallel with the external resistor. Regardless of whether acid or alkali pollution is present, continuously collect voltage data and send it to the monitoring terminal. In the absence of acid or alkali pollution, the collected and displayed voltage signal is the system baseline voltage. When acid or alkali pollution enters the water and comes into contact with the cathode, the data acquisition unit collects the voltage data across the external resistor and sends it to the monitoring terminal. Acid pollution will cause the voltage to rise rapidly, while pollution reduction will cause the voltage to drop rapidly. Step 3: In the monitoring terminal, a water pH identification model based on voltage data is constructed through machine learning. The voltage data is converted into water pH and displayed online in real time, thereby realizing the identification and prediction of water acid and alkali pollution.

[0012] Furthermore, aquatic plants are planted in the soil of the submerged tank.

[0013] In this invention, a water pH identification model based on voltage data is constructed using machine learning. This model converts the collected voltage signals into corresponding water pH values, enabling automatic identification and quantitative analysis of acid and alkali pollution. The machine learning identification method includes the following sub-steps: Step 1, Data Preprocessing and Feature Extraction: The raw voltage sequence acquired by the data acquisition device undergoes time-series smoothing and noise filtering to extract multi-dimensional feature parameters, including steady-state voltage value, short-time rate of change, peak response amplitude, gradient features, integral area, and frequency domain power spectrum features, forming a feature vector set. These features can be single-point time-series features or statistical features within a sliding window, thus ensuring the model's sensitivity to transient pollution events.

[0014] Step 2, Model Construction and Training: Based on the feature vectors and corresponding pH calibration data, a supervised learning method is used to train the recognition model. The model includes, but is not limited to, any one or more combinations of fully connected neural networks, convolutional neural networks, recurrent neural networks, support vector machines, random forests, or gradient boosting trees. The model uses a nonlinear mapping function to realize the mapping relationship between voltage features and water pH.

[0015] Step 3, Model Inference and Output: During system operation, the monitoring terminal receives voltage data uploaded by the sensors in real time, executes the feature extraction and model inference modules, and outputs the predicted pH value, which is dynamically displayed in numerical or curve form. The system also automatically identifies abnormal pollution events based on the voltage mutation rate, the predicted pH change gradient, and the duration threshold.

[0016] In a specific embodiment of the present invention, the specific steps of the machine learning recognition method are as follows: 1. Voltage signal acquisition and preprocessing The data acquisition unit collects the output voltage of the floating island microbial fuel cell in real time at a sampling rate of once per second, and then processes it through filtering, noise reduction, and smoothing. The system extracts multi-dimensional feature parameters, including steady-state voltage, short-time rate of change, maximum response amplitude, signal gradient, integral area, and power spectral density characteristics.

[0017] The above features are used to construct the input feature vector using a sliding window approach.

[0018] 2. Model Training Experimental data under different pH conditions (pH=2~13) were collected as a sample set, and a voltage feature-pH mapping model was established using supervised learning. The model adopts a convolutional neural network structure, with the aforementioned feature vector as the input layer and the predicted pH value as the output layer.

[0019] 3. Model Inference and Real-time Display After receiving voltage data, the monitoring terminal executes a model inference process and displays the predicted pH results in real time as numerical values ​​and dynamic curves. When a sudden change occurs in the predicted pH, the system identifies it as a pollution event and triggers an early warning mechanism.

[0020] The floating island-type microbial fuel cell sensor provided by this invention buries the anode in the soil of a submerged tank, which is suspended below the floating island. This eliminates the need to submerge the anode at the bottom of the wetland, making it suitable for monitoring deep wetlands. The device carries a data acquisition unit and a solar panel, enabling real-time online transmission of electrical signal data to computers and mobile terminals. By introducing machine learning algorithms, a nonlinear mapping model between the output voltage signal of the microbial fuel cell and the pH value of the water is established, allowing the monitoring terminal to intelligently analyze the real-time voltage data, thereby achieving high-precision, continuous, and online monitoring of the water pH value. Simultaneously, planting aquatic plants provides a continuous carbon source for the soil-generating bacteria, which helps to increase the voltage of the microbial fuel cell, maintain a stable voltage output over a long period, and prevent the limited soil organic matter in the submerged tank from being depleted, thus maintaining power generation. Furthermore, the aquatic plants beautify the sensor's appearance and create a scenic effect. Attached Figure Description

[0021] Figure 1 This is a schematic diagram of the floating island-type microbial fuel cell sensor of the present invention. Wherein: 1-floating island, 2-submerged tank, 3-anode, 4-cathode, 5-wire, 6-external resistance, 7-data acquisition unit, 8-solar panel, 9-support frame, and 10-monitoring terminal.

[0022] Figure 2 The figure shows the alkaline contamination voltage signal measured by the floating island-type microbial fuel cell sensor of the present invention in Example 1. The number on each voltage peak represents the pH value of the alkaline contamination.

[0023] Figure 3 The figure shows the acid pollution voltage signal measured by the floating island-type microbial fuel cell sensor of the present invention in Example 2. The number on each voltage peak represents the pH value of the acid pollution. Detailed Implementation

[0024] The preferred embodiments of the present invention will now be described in detail with reference to specific examples. It should be understood that the following examples are given for illustrative purposes only and are not intended to limit the scope of the invention. Those skilled in the art can make various modifications and substitutions to the present invention without departing from its spirit and essence.

[0025] Unless otherwise specified, the experimental methods used in the following examples are conventional methods.

[0026] Unless otherwise specified, all materials and reagents used in the following examples are commercially available. Example 1

[0027] A floating island-type microbial fuel cell sensor was constructed in the center of a wetland approximately 3 meters deep on the Xianlin campus of Nanjing Normal University. For example... Figure 1 As shown, the floating island (1) is a square foam with a side length of 40 cm and a thickness of 10 cm. A 10-mesh stainless steel mesh with a side length of 40 cm is fixed on the bottom surface of the foam as the cathode (4). Then, the submerged bucket (2) is embedded below the opening in the center of the floating island. The soil in the submerged bucket is planted with aquatic plants such as Alisma plantago-aquatica. A square carbon felt (10 cm on each side) is completely buried in the bottom of the soil in the submerged bucket as the anode (3), and is connected in series with the anode and a 20 kΩ external resistor (6) through a titanium wire (5). Above the floating island, a 12V solar panel (8) is used to power the data acquisition device (7). The data acquisition device (7) is connected in parallel with the external resistor (6) and records and remotely sends a voltage data to the monitoring terminal (10) every 1 second.

[0028] Subsequently, 100 mL of buffer solutions with pH 10.6, pH 11.2, pH 11.8, pH 12.4, and pH 13.0 were added to the water every 20 minutes, producing voltage peaks of -6.4 mV, -13.8 mV, -16.6 mV, -17.8 mV, and -22.6 mV, respectively. Figure 2 Based on the prediction results of the deep learning model, the pH values ​​displayed by the monitoring terminal were 10.69, 11.12, 11.81, 12.44, and 12.96. Example 2

[0029] A floating island-type microbial fuel cell sensor was constructed in the center of a wetland approximately 3 meters deep at the Xianlin Campus of Nanjing Normal University. The floating island was a square foam with sides of 40 cm and a thickness of 10 cm. A 10-mesh stainless steel mesh with sides of 40 cm was fixed to the bottom of the foam as the cathode. A submerged tank was then embedded below the central opening of the floating island. A stainless steel pipe (φ5.0 cm × 10 cm) was inserted completely into the soil of the submerged tank as the anode and connected in series with a 20 kΩ external resistor via a titanium wire. A 12 V solar panel powered the data acquisition device above the floating island. The data acquisition device was connected in parallel with the external resistor and recorded and remotely transmitted one voltage data point every second. It was found that without any plants, the baseline voltage was 20 mV, lower than the baseline voltage of 100 mV in Example 1.

[0030] 100 mL of buffer solutions with pH 4.6, pH 4.2, pH 3.4, pH 2.8, and pH 2.2 were added to the water sequentially every 10 minutes, producing voltage peaks of 2.2 mV, 3.1 mV, 4.5 mV, 6.2 mV, and 7.5 mV, respectively. Figure 3 As shown in the figure, the pH values ​​displayed by the monitoring terminal based on the prediction results of the deep learning model were 4.72, 4.12, 3.39, 2.87, and 2.35, which are close to the pH of the added buffer solution. This indicates that the floating island-type microbial fuel cell sensor can collect water changes in real time and achieve high-precision, continuous, and online monitoring through machine learning models.

Claims

1. A floating island-type microbial fuel cell sensor, characterized in that, Includes a submersible tank, cathode, external resistor, data acquisition unit, solar panel, and monitoring terminal; The submersible tank is positioned below the floating island and submerged in the water; the floating island floats on the water surface; the support frame is positioned on the surface of the floating island, the external resistor and data acquisition device are fixed on the support frame, and the solar panel is positioned on top of the support frame. The submerged tank contains soil, and the anode is buried in the soil; the cathode is fixed to the bottom of the floating island and submerged in the water; the anode and cathode assembly are connected in series with an external resistor via wires to form a microbial fuel cell; the external resistor is connected in parallel with a data acquisition device, and the data acquisition device transmits the voltage data across the external resistor of the microbial fuel cell to a monitoring terminal.

2. The floating island-type microbial fuel cell sensor according to claim 1, characterized in that, The anode is made of stainless steel, graphite, or conductive carbon fiber.

3. The floating island-type microbial fuel cell sensor according to claim 1, characterized in that, The cathode is made of stainless steel or titanium and is in the shape of a mesh or sheet.

4. The floating island-type microbial fuel cell sensor according to claim 1, characterized in that, The conductor is made of titanium.

5. The floating island-type microbial fuel cell sensor according to claim 1, characterized in that, The floating island has a through-hole in the middle for embedding a submersible bucket.

6. The floating island-type microbial fuel cell sensor according to claim 1, characterized in that, Aquatic plants are planted in the soil of the submerged tank.

7. A method for online monitoring of water pollution events using a floating island-type microbial fuel cell sensor as described in any one of claims 1-6, characterized in that, Includes the following steps: Step 1: Float the floating island on the water surface, immerse the cathode below the water surface, immerse the submersible bucket suspended below the floating island completely in the water, bury the anode in the soil of the submersible bucket, and fix the external resistor, data acquisition device and solar panel above the floating island. Step 2: Connect the anode and cathode in series with the external resistor via wires to form a microbial fuel cell. Connect the data acquisition device in parallel with the external resistor to continuously collect voltage data and send it to the monitoring terminal. In the absence of acid or alkali pollution, the collected voltage signal is the baseline voltage of the system. When acid or alkali pollution enters the water and comes into contact with the cathode, the data acquisition device collects the voltage data across the external resistor and sends it to the monitoring terminal. Step 3: In the monitoring terminal, a water pH identification model based on voltage data is constructed through machine learning. The voltage data is converted into water pH and displayed online in real time, thereby realizing the identification and prediction of water acid and alkali pollution.

8. The method according to claim 7, characterized in that, Aquatic plants are planted in the soil of the submerged tank.

Images