Dynamic monitoring and intelligent adjusting method of ship ballast water treatment system
By combining spectral detection and water quality assessment models with reinforcement learning algorithms, the concentration of microorganisms in ship ballast water is dynamically monitored, and intelligent adjustment and treatment schemes are implemented. This solves the problem of insufficient adaptive adjustment in existing systems and achieves energy-saving and efficient ballast water treatment.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-11-14
- Publication Date
- 2026-03-31
AI Technical Summary
Existing ship ballast water treatment systems lack adaptive adjustment capabilities, leading to increased energy consumption, unstable treatment effects, and difficulty in meeting international environmental standards and economic requirements.
By combining spectral detection with a water quality assessment model and reinforcement learning algorithm, the concentration of microorganisms in ballast water is dynamically monitored and the treatment scheme is intelligently adjusted, including spectral acquisition, water quality status determination, and generation of optimized treatment schemes.
It achieves flexibility and adaptability in ballast water treatment, reduces energy consumption, ensures that the treatment effect meets international environmental standards, reduces operating costs, and improves treatment efficiency.
Smart Images
Figure CN121762473A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of ship ballast water treatment technology, specifically to a method for dynamic monitoring and intelligent adjustment of ship ballast water treatment systems. Background Technology
[0002] With the implementation of the International Maritime Organization's Convention on the Control and Management of Ballast Water and Sediments, ballast water, as a crucial medium for maintaining stability and safety during ship navigation, has become a key issue for global marine environmental protection. Existing ballast water treatment technologies mainly include physical methods such as filtration and ultraviolet irradiation, chemical methods such as ozone and chlorine dosing, and combined methods such as electrolysis and ultrasound. Various devices have been developed for engineering applications, effectively eliminating or killing invasive aquatic organisms and harmful microorganisms in ballast water. Most mainstream ships are now equipped with ballast water treatment systems, which typically collect water quality parameters through sensors, and the ship's management system records the data and verifies compliance. Some advanced systems also integrate remote monitoring and data uploading functions, enabling the exchange of operational data between ship and shore, providing reference for ship operations and regulatory agencies.
[0003] However, traditional ballast water treatment systems mostly operate in fixed modes, lacking the ability to adapt to changes in water quality and environmental conditions. This mode easily leads to increased energy consumption, and in some cases, even under- or over-treatment, affecting both treatment effectiveness and increasing operating costs. During long voyages or in complex sea conditions, water quality parameters often fluctuate significantly. If control is still based on preset parameters, it is difficult to guarantee long-term stable compliance of ballast water discharge with international standards. When dealing with the differentiated regulations of various ports, existing systems are insufficient in terms of dynamism and intelligence, failing to simultaneously meet the dual requirements of environmental compliance and ship economics. Summary of the Invention
[0004] To address the shortcomings of existing technologies, this invention provides a dynamic monitoring and intelligent adjustment method for ship ballast water treatment systems. The technical problem this invention aims to solve is: how to dynamically monitor the concentration of microorganisms in ship ballast water and intelligently adjust the treatment scheme by combining spectral detection with a water quality evaluation model and reinforcement learning algorithms, thereby achieving energy-efficient and high-performance ballast water treatment that meets international environmental standards.
[0005] To achieve the above objectives, the present invention is implemented through the following technical solution: a dynamic monitoring and intelligent adjustment method for a ship ballast water treatment system, comprising: S1. acquiring spectral feature data by performing spectral acquisition on the ballast water flow entering the ballast water treatment device through a spectral detection unit, and generating spectral monitoring data based on the spectral feature data to identify the types and concentrations of microorganisms in the ballast water flow.
[0006] S2. Input the spectral monitoring data into the water quality evaluation model, and calculate the water quality status data of the microbial content in the ballast water flow through the water quality evaluation model. The water quality evaluation model includes preset thresholds and judgment rules.
[0007] S3. Input the water quality status data into the ballast water treatment control unit, which generates an optimized treatment scheme based on the spectral feature data using a reinforcement learning algorithm.
[0008] S4. The optimized processing plan is sent to the ballast water treatment device for execution and operational feedback data is collected.
[0009] S5. Input the optimized processing scheme and the operation feedback data into the ship energy consumption management system to generate a compliance data package, and output the compliance data package to the ship management terminal for verification and archiving.
[0010] Preferably, the spectral detection unit uses a combined wavelength band for spectral acquisition, wherein the combined wavelength band has a spectral resolution ≥1 nm and a sampling time interval ≤100 nm. The combined band includes the ultraviolet band and the near-infrared band, with the effective ultraviolet band having a wavelength range of 200nm-380nm and the effective near-infrared band having a wavelength range of 700nm-1100nm.
[0011] Preferably, the water quality status data includes a compliant status and a pending status. The water quality status data is determined based on the preset threshold, which includes a microbial concentration threshold. The microbial concentration threshold is 10 when the minimum size of the microorganism is 50 micrometers. When the minimum size of the microorganism is between 10 micrometers and 50 micrometers, the concentration threshold of the microorganism is 10. The determination rule is that when the microbial concentration is lower than the microbial concentration threshold, the water quality is determined to be in a qualified state; when the microbial concentration is higher than the microbial concentration threshold, the water quality is determined to be in a state to be treated.
[0012] Preferably, the ballast water treatment control unit includes an ultraviolet irradiation module, a chemical dosing module, and a filtration module, wherein the irradiation intensity adjustment range of the ultraviolet irradiation module is 100. -400 The dosage adjustment range of the chemical dosing module is 1. -5 The filtration accuracy adjustment range of the filter module is 0.1. -5 .
[0013] Preferably, the reinforcement learning algorithm generates the optimized processing scheme by adjusting the irradiation intensity, the dosage, and the filtering accuracy.
[0014] Preferably, the reinforcement learning algorithm updates the optimization scheme by calculating policy parameters, and the model formula of the reinforcement learning algorithm is: in, These are strategy parameters, dimensionless. The learning rate is dimensionless. For the policy function, The input state after normalization is dimensionless. The output action vector has a value range of [0,1] and is dimensionless. For policy gradient, The reward value is dimensionless.
[0015] Preferably, the operational feedback data includes ultraviolet irradiation dose, online concentration of chemical agents, filtration pressure difference, and instantaneous flow rate.
[0016] Preferably, the compliance data packet is output to the ship management terminal in the International Maritime Organization standard format.
[0017] This invention provides a method for dynamic monitoring and intelligent adjustment of a ship's ballast water treatment system. It has the following beneficial effects: This invention combines spectral detection and a water quality assessment model to monitor the types and concentrations of microorganisms in ship ballast water in real time, accurately determine the water quality status, and avoid the fixed parameter settings of traditional systems. This improves the flexibility and adaptability of water treatment and ensures effective operation under different environmental conditions.
[0018] The dynamic monitoring and intelligent adjustment method of the ship ballast water treatment system uses reinforcement learning algorithm to optimize and adjust the ballast water treatment process, intelligently generate treatment plans and adjust them according to real-time feedback data, which significantly reduces energy consumption and ensures that the ship's ballast water discharge meets international environmental protection standards, improves the system's energy efficiency and compliance, reduces operating costs and improves treatment efficiency. Attached Figure Description
[0019] Figure 1 This is a schematic diagram of a ship's ballast water treatment system. Figure 2 This is a schematic diagram of the dynamic monitoring and intelligent adjustment process; Figure 3 This is a schematic diagram of the operation of the spectral detection unit and the water quality assessment model; Figure 4 This is a schematic diagram of data processing for ship energy consumption management and compliance. Detailed Implementation
[0020] The technical solutions of 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. Example 1
[0021] like Figure 1-4 As shown, this embodiment of the invention provides a dynamic monitoring and intelligent adjustment method for a ship ballast water treatment system, including: S1. Obtaining spectral feature data by acquiring the spectral characteristics of the ballast water flow entering the ballast water treatment device through a spectral detection unit, and generating spectral monitoring data based on the spectral feature data to identify the types and concentrations of microorganisms in the ballast water flow. The spectral detection unit uses a combined band for spectral acquisition, with a combined band spectral resolution ≥1nm and a sampling time interval ≤100 nm. The combined band includes the ultraviolet band and the near-infrared band. The effective wavelength of the ultraviolet band is 200nm, and the effective wavelength of the near-infrared band is 700nm.
[0022] S2. Input the spectral monitoring data into the water quality assessment model. The model calculates the water quality status data of microbial content in the ballast water flow. The water quality assessment model includes preset thresholds and judgment rules. The water quality status data includes compliant and untreated states. The water quality status data is judged based on preset thresholds, including a microbial concentration threshold. When the minimum size of microorganisms is 10 micrometers, the microbial concentration threshold is 10. The judgment rule is as follows: when the microbial concentration is below the microbial concentration threshold, the water quality is judged to be in a compliant state; when the microbial concentration is above the microbial concentration threshold, the water quality is judged to be in a state requiring treatment. The spectral detection unit obtains the spectral characteristic data of the ballast water flow through spectral acquisition, and then identifies the types and concentrations of microorganisms in the water flow. The system inputs the size and concentration values of the microorganisms into the water quality evaluation model, and judges the water quality status according to the preset threshold and judgment rule.
[0023] Setting the microbial concentration threshold: Based on the rules set in the water quality assessment model, the system compares the minimum size of microorganisms with the microbial concentration threshold and determines the water quality status: When the minimum size of the microorganism is 10 micrometers, the microbial concentration threshold is set to 10. .
[0024] According to the preset rules: If the microbial concentration is below the threshold, the water quality is considered to be up to standard.
[0025] If the microbial concentration is higher than the threshold, the water quality is determined to be in a state of needing treatment.
[0026] Determine water quality status: In practice, the system compares the real-time collected microbial concentration data with the above thresholds to determine the water quality status. Assuming the measured microbial concentration is 8 / The system will determine that the water quality meets the standards and no further treatment is required.
[0027] Assuming the measured microbial concentration is 12 / The system will determine that the water quality is in a state of needing treatment and will start the ballast water treatment device to process it.
[0028] S3. Input the water quality status data into the ballast water treatment control unit. The ballast water treatment control unit generates an optimized treatment plan based on the spectral feature data using a reinforcement learning algorithm. The ballast water treatment control unit includes an ultraviolet irradiation module, a chemical dosing module, and a filtration module. The irradiation intensity of the ultraviolet irradiation module is 100. The chemical dosing module has a dosing capacity of 1. The filtration accuracy of the filter module is 0.1. Reinforcement learning algorithms generate optimized treatment schemes by adjusting illumination intensity, dosage, and filtering accuracy. The algorithm updates the optimized scheme by calculating policy parameters. The model formula for reinforcement learning is: in, These are strategy parameters, dimensionless. The learning rate is dimensionless. For the policy function, The input state after normalization is dimensionless. This outputs an action vector, with a value of 0, which is dimensionless. For policy gradient, The reward value is dimensionless.
[0029] Assuming initial policy parameters The learning rate η is 0.05, and the input state is 0.5. The value is 10, and the action is selected through the policy function. The value is 0. The gradient update of the system will affect the subsequent policy parameters.
[0030] After several iterations, the system found that selecting action 0 was not conducive to the final water treatment effect, so the strategy parameters will be adjusted so that the system can select other non-zero actions.
[0031] S4. The optimized treatment plan is sent to the ballast water treatment unit for execution, and operational feedback data is collected. Operational feedback data includes UV radiation dose, online chemical concentration, filtration differential pressure, and instantaneous flow rate.
[0032] S5. Input the optimized processing plan and operational feedback data into the ship energy consumption management system to generate a compliance data package, and output the compliance data package to the ship management terminal for verification and archiving. The compliance data package is output to the ship management terminal in the International Maritime Organization standard format.
[0033] When the parameters are at their minimum values, energy and resource consumption are minimized, resulting in lower operating costs. The system operates under light load, making it suitable for situations with good water quality or where excessive treatment is not required. This improves the system's economic efficiency and makes it suitable for long-term operation.
[0034] Example 2 Based on Embodiment 1, this embodiment, through its implementation method of maximizing configuration, can achieve the best ballast water treatment effect in highly polluted waters or under special ship operating conditions.
[0035] S1. The spectral detection unit acquires spectral characteristic data of the ballast water flow entering the ballast water treatment device. Based on the spectral characteristic data, the types and concentrations of microorganisms in the ballast water flow are identified, generating spectral monitoring data. The spectral detection unit uses a combined wavelength range for spectral acquisition, with a combined wavelength range spectral resolution ≥1nm and a sampling time interval ≤100 nm. The combined band includes the ultraviolet band and the near-infrared band. The effective wavelength of the ultraviolet band is 380nm, and the effective wavelength of the near-infrared band is 1100nm.
[0036] S2. Input the spectral monitoring data into the water quality assessment model. The model calculates the water quality status data of microbial content in the ballast water flow. The water quality assessment model includes preset thresholds and judgment rules. The water quality status data includes compliant and untreated states. The water quality status data is judged based on preset thresholds, including a microbial concentration threshold. The microbial concentration threshold is set to 10 when the minimum size of the microorganisms is 50 micrometers. The judgment rule is that when the microbial concentration is lower than the microbial concentration threshold, the water quality is judged to be in a qualified state; when the microbial concentration is higher than the microbial concentration threshold, the water quality is judged to be in a state to be treated.
[0037] Assuming the measured microbial concentration is 8 The system will determine that the water quality meets the standards and no further treatment is required.
[0038] Assuming the measured microbial concentration is 12 The system will determine that the water quality is in a state of needing treatment and will start the ballast water treatment device to process it.
[0039] S3. Input the water quality status data into the ballast water treatment control unit. The ballast water treatment control unit generates an optimized treatment plan based on the spectral feature data using a reinforcement learning algorithm. The ballast water treatment control unit includes an ultraviolet irradiation module, a chemical dosing module, and a filtration module. The irradiation intensity of the ultraviolet irradiation module is 400. The chemical dosing module has a dosing capacity of 5. The filtration accuracy of the filter module is 5. Reinforcement learning algorithms generate optimized treatment schemes by adjusting illumination intensity, dosage, and filtering accuracy. The algorithm updates the optimized scheme by calculating policy parameters. The model formula for reinforcement learning is: in, These are strategy parameters, dimensionless. The learning rate is dimensionless. For the policy function, The input state after normalization is dimensionless. This is the output action vector, with a value of 1, and is dimensionless. For policy gradient, The reward value is dimensionless.
[0040] Assume the initial policy parameters are =0.5, learning rate =0.05, output action vector =1, and calculate the policy gradient as 0.7, reward value =5.
[0041] Updated formula: The updated strategy parameter is 0.535, which makes the system's selected action 1 more optimized in subsequent operations.
[0042] Iterate again: The system will adjust according to the new strategy parameters. Further operational adjustments will be made to ensure that the optimal operational strategy is selected under different conditions.
[0043] S4. The optimized treatment plan is sent to the ballast water treatment unit for execution, and operational feedback data is collected. Operational feedback data includes UV radiation dose, online chemical concentration, filtration differential pressure, and instantaneous flow rate.
[0044] S5. Input the optimized processing plan and operational feedback data into the ship energy consumption management system to generate a compliance data package, and output the compliance data package to the ship management terminal for verification and archiving. The compliance data package is output to the ship management terminal in the International Maritime Organization standard format.
[0045] When the parameters are set to their maximum values, the strongest treatment effect is ensured, enabling effective water treatment even in extreme environments. This guarantees that the water quality meets international environmental standards, avoiding undertreatment issues. It provides the highest level of safety assurance and is suitable for situations with significant water quality fluctuations.
[0046] Example 3 Based on the above embodiments, this embodiment selects intermediate parameters within the technical scope of the present invention to maintain processing efficiency and effectiveness while taking into account energy saving and operating costs, providing a highly efficient, economical solution for ballast water treatment of ships that meets international environmental protection requirements.
[0047] S1. The spectral detection unit acquires spectral characteristic data of the ballast water flow entering the ballast water treatment device. Based on the spectral characteristic data, the types and concentrations of microorganisms in the ballast water flow are identified, generating spectral monitoring data. The spectral detection unit uses a combined wavelength range for spectral acquisition, with a combined wavelength range spectral resolution ≥1nm and a sampling time interval ≤100 nm. The combined band includes the ultraviolet band and the near-infrared band. The effective wavelength of the ultraviolet band is 290nm, and the effective wavelength of the near-infrared band is 900nm.
[0048] S2. Input the spectral monitoring data into the water quality assessment model. The model calculates the water quality status data of microbial content in the ballast water flow. The water quality assessment model includes preset thresholds and judgment rules. The water quality status data includes compliant and untreated states. The water quality status data is judged based on preset thresholds, including a microbial concentration threshold. When the minimum size of microorganisms is 30 micrometers, the microbial concentration threshold is 10. The judgment rule is that when the microbial concentration is lower than the microbial concentration threshold, the water quality is judged to be in a qualified state; when the microbial concentration is higher than the microbial concentration threshold, the water quality is judged to be in a state to be treated.
[0049] Setting the microbial concentration threshold: Based on the rules set in the water quality assessment model, the system compares the minimum size of microorganisms with the microbial concentration threshold and determines the water quality status: When the minimum size of the microorganism is 30 micrometers, the microbial concentration threshold is set to 10. .
[0050] According to the preset rules: If the microbial concentration is below the threshold, the water quality is considered to be up to standard.
[0051] If the microbial concentration is higher than the threshold, the water quality is determined to be in a state of needing treatment.
[0052] Determine water quality status: In practice, the system compares the real-time collected microbial concentration data with the above thresholds to determine the water quality status. Assuming the measured microbial concentration is 8 / The system will determine that the water quality meets the standards and no further treatment is required.
[0053] Assuming the measured microbial concentration is 12 / The system will determine that the water quality is in a state of needing treatment and will start the ballast water treatment device to process it.
[0054] S3. Input the water quality status data into the ballast water treatment control unit. The ballast water treatment control unit generates an optimized treatment plan based on the spectral feature data using a reinforcement learning algorithm. The ballast water treatment control unit includes an ultraviolet irradiation module, a chemical dosing module, and a filtration module. The irradiation intensity of the ultraviolet irradiation module is 250. The chemical dosing module has a dosing capacity of 3. The filtration accuracy of the filter module is 2.5. Reinforcement learning algorithms generate optimized treatment schemes by adjusting illumination intensity, dosage, and filtering accuracy. The algorithm updates the optimized scheme by calculating policy parameters. The model formula for reinforcement learning is: in, These are strategy parameters, dimensionless. The learning rate is dimensionless. For the policy function, The input state after normalization is dimensionless. The output action vector has a value of 0.5 and is dimensionless. For policy gradient, The reward value is dimensionless.
[0055] Assume the initial policy parameters are =0.5, learning rate =0.05, output action vector =0.5, and calculate the policy gradient as 0.6, reward value =4.
[0056] Updated formula: Adjustments to control parameters: Based on the updated strategy, control parameters such as UV irradiance intensity and chemical dosage will be further optimized. The initial UV irradiance intensity was 250 mJ / cm², and the new UV irradiance intensity will be 132.5 mJ / cm².
[0057] Continuous iteration: Through multiple iterations and learning, the system will continue to adjust strategy parameters and optimize the water treatment process, ultimately achieving the best treatment effect and energy efficiency.
[0058] S4. The optimized treatment plan is sent to the ballast water treatment unit for execution, and operational feedback data is collected. Operational feedback data includes UV radiation dose, online chemical concentration, filtration differential pressure, and instantaneous flow rate.
[0059] S5. Input the optimized processing plan and operational feedback data into the ship energy consumption management system to generate a compliance data package, and output the compliance data package to the ship management terminal for verification and archiving. The compliance data package is output to the ship management terminal in the International Maritime Organization standard format.
[0060] By averaging the selected parameters, the system effectively balances energy consumption and water quality treatment stability while ensuring treatment effectiveness. It is suitable for most routine situations, ensuring an optimal balance between treatment results and resource consumption. This guarantees system stability and continuity, avoiding unnecessary waste caused by extreme treatment methods.
[0061] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.
Claims
1. A method for dynamic monitoring and intelligent adjustment of a ship ballast water treatment system, characterized in that, The method comprises the following steps: S1. Obtain spectral feature data by performing spectral acquisition on the ballast water flowing into the ballast water treatment device through a spectral detection unit, identify the species and concentration of microorganisms in the ballast water based on the spectral feature data, and generate spectral monitoring data; S2. Input the spectral monitoring data into a water quality evaluation model, calculate the water quality state data of the microorganism content in the ballast water through the water quality evaluation model, and the water quality evaluation model comprises a preset threshold and a judgment rule; S3. Input the water quality state data into a ballast water treatment control unit, and generate an optimized treatment scheme through a reinforcement learning algorithm according to the spectral feature data; S4. Execute the optimized treatment scheme in the ballast water treatment device and collect operation feedback data; S5. Input the optimized treatment scheme and the operation feedback data into a ship energy consumption management system to generate compliance data packets, and output the compliance data packets to a ship management terminal for verification and archiving.
2. The method of dynamic monitoring and intelligent adjustment of a ship's ballast water treatment system according to claim 1, characterized in that: The spectral detection unit adopts a joint wave band for the spectral acquisition, the joint wave band has a spectral resolution of ≥1nm and a sampling time interval of ≤100 , the joint wave band includes an ultraviolet wave band and a near-infrared wave band, the effective wave band of the ultraviolet wave band is 200nm-380nm, and the effective wave band of the near-infrared wave band is 700nm-1100nm.
3. The method of dynamic monitoring and intelligent adjustment of a ship's ballast water treatment system according to claim 1, characterized in that: The water quality state data includes a qualified state and a to-be-treated state, the water quality state data is determined based on the preset threshold, the preset threshold includes a microorganism concentration threshold, when the minimum size of the microorganism is 50 microns, the microorganism concentration threshold is 10 , when the minimum size of the microorganism is between 10 microns and 50 microns, the microorganism concentration threshold is 10 , and the determination rule is that when the microorganism concentration is lower than the microorganism concentration threshold, the water quality state is determined as the qualified state, and when the microorganism concentration is higher than the microorganism concentration threshold, the water quality state is determined as the to-be-treated state.
4. The method for dynamic monitoring and intelligent adjustment of a ship ballast water treatment system according to claim 1, characterized in that: The ballast water treatment control unit comprises a UV irradiation module, a chemical dosing module and a filtration module, the irradiation intensity of the UV irradiation module is adjusted in the range of 100 -400 , the dosing amount of the chemical dosing module is adjusted in the range of 1 -5 , and the filtration precision of the filtration module is adjusted in the range of 0.1 -5 .
5. The method for dynamic monitoring and intelligent adjustment of ship ballast water treatment system according to claim 4, characterized in that: The reinforcement learning algorithm generates the optimized treatment scheme by adjusting the irradiation intensity, the dosage and the filtration accuracy.
6. The method for dynamic monitoring and intelligent adjustment of ship ballast water treatment system according to claim 1, characterized in that: The reinforcement learning algorithm updates the optimized treatment scheme by calculating the policy parameter, and the model formula of the reinforcement learning algorithm is: ， wherein, is a policy parameter, is a learning rate, is a policy function, is a normalized input state, is an output action vector, with a value range of [0, 1], is a policy gradient, and is a reward value.
7. The method for dynamic monitoring and intelligent adjustment of ship ballast water treatment system according to claim 1, characterized in that: The operation feedback data includes ultraviolet irradiation dose, online concentration of chemical agent, filtration pressure difference and instantaneous flow.
8. The method for dynamic monitoring and intelligent adjustment of ship ballast water treatment system according to claim 1, characterized in that: The compliance data packets are output to the ship management terminal in the format of the International Maritime Organization standard.