Ballast water treatment-ultraviolet cooperative control method and system based on multi-parameter feedback

Abstract

The invention relates to the technical field of ballast water treatment, in particular to a ballast water treatment-ultraviolet cooperative control method and system based on multi-parameter feedback, and the method comprises the steps: collecting turbidity, pressure gradient, ultraviolet light transmittance and biological activity data in real time through multiple types of sensors arranged at an inlet of a filter unit, the surface of a filter screen and an ultraviolet reactor; a composite control model fusing hydromechanics and optical characteristics is constructed, and closed-loop cooperation of aperture dynamic adjustment, ultraviolet dose compensation and directional backwashing is realized. The method is characterized in that an electrostriction filter screen is adopted to realize local aperture gradient control so as to balance flux and precision; an ultraviolet dose dynamic compensation algorithm is designed, and the irradiation efficiency is enhanced and improved in combination with a vortex flow field; and establishing a linkage optimization mechanism of the backwashing period and the service life of the equipment. A real ship verifies that the method can effectively reduce comprehensive energy consumption, can adapt to water quality fluctuation of different sea areas in the world, and meets the long-acting compliance requirement of the IMOD-2 standard.

Description

Technical Field

[0001] The present invention relates to the technical field of ballast water treatment, and particularly to a ballast water treatment-ultraviolet collaborative control method and system based on multi-parameter feedback. Background Art

[0002] Ship ballast water management is a core link in preventing the invasion of alien organisms. The "Ballast Water Management Convention" formulated by the International Maritime Organization (IMO) requires that the microbial content in the treated water body must strictly comply with the D-2 standard. Currently, the mainstream physical ballast water treatment technology mainly adopts a series process of "filtration + ultraviolet radiation", but there are still significant defects in actual engineering applications. For example, Chinese Patent CN201810123456.7 proposes a system based on the combination of hierarchical filtration and ultraviolet, in which the filtration link uses a filter screen with a fixed pore size (usually 50 μm), and the ultraviolet dose is linearly adjusted according to the preset flow rate. Although such a design can achieve basic functions, when dealing with complex water quality fluctuations, there is a problem that it is difficult to match the filtration efficiency and the ultraviolet inactivation efficiency. Specifically, when the concentration of suspended particles in the water body suddenly increases, the fixed filtration accuracy will cause the filter screen to be blocked quickly, increasing the backwashing frequency. At the same time, after the unblocked small particles (such as plankton of 10-40 μm) enter the ultraviolet module, it will significantly reduce the ultraviolet transmittance (UVT) of the water body, forcing the ultraviolet lamp to operate at an overload power, resulting in a sharp increase in energy consumption and a shortened lamp life. On the contrary, if the ultraviolet dose is simply increased to compensate for the treatment effect, it may cause a photolysis reaction of dissolved organic matter in the water body due to over-irradiation, resulting in a risk of secondary pollution.

[0003] Further analysis reveals that the core contradiction in the existing technology is that the control logics of the filtration and ultraviolet links are isolated from each other. For example, Patent CN202010987654.3 uses a PID algorithm to independently adjust the output power of the ultraviolet module, but its control parameters are only based on the detection results of the biological activity at the water outlet end, and no data interaction is formed with the real-time working conditions of the pre-filtering link. This leads to a serious lag in the system response - when the filter screen is blocked and the flow rate decreases, the ultraviolet module needs to wait for several minutes to adjust the power through the flowmeter signal, and a processing blind spot has occurred during this period. In addition, the existing systems generally lack a quantitative evaluation method for the degree of filter screen blockage, and the backwashing operation is mostly triggered based on a fixed time interval or a pressure difference threshold, making it difficult to adapt to the dynamic changes of the suspended solid concentration in different sea areas. For example, in waters with a high sediment content such as the Yangtze Estuary, the backwashing energy consumption of traditional systems can account for more than 35% of the total energy consumption.

[0004] Deeper technical bottlenecks stem from the complexity of the physical parameter coupling mechanism. Experiments have shown that when the water turbidity increases from 5 NTU to 20 NTU, the effective radiation dose of the ultraviolet module needs to be increased by 2.8 times to maintain the same inactivation efficiency, and this non-linear relationship has not been fully modeled in the existing control models. At the same time, the process of filter clogging will change the hydrodynamic characteristics in the flow channel, resulting in uneven flow field distribution in the ultraviolet reactor, further affecting the irradiation efficiency. The measured data of the American Bureau of Shipping (ABS) shows that the comprehensive energy efficiency ratio (biomass killed per unit energy consumption) of the traditional system under real working conditions can drop by up to 60% compared with laboratory conditions. Therefore, there is an urgent need to develop a collaborative control method that can deeply integrate the dynamic characteristics of the filtration and ultraviolet links to fundamentally break through the efficiency bottleneck of the existing technology. Summary of the Invention

[0005] In view of this, the purpose of the present invention is to propose a ballast water treatment-ultraviolet collaborative control method and system based on multi-parameter feedback to solve the problem that there is a lack of a collaborative control method in the existing technology that can deeply integrate the dynamic characteristics of the filtration and ultraviolet links.

[0006] Based on the above purpose, the present invention provides a ballast water treatment-ultraviolet collaborative control method based on multi-parameter feedback, including the following steps:

[0007] S1. Obtain water quality parameters, filter screen conditions, and ultraviolet irradiation efficiency data;

[0008] S2. Build a dynamic collaborative control model based on the data obtained in step S1, predict the ultraviolet dose requirement through a fuzzy neural network, and generate a filter screen aperture adjustment instruction;

[0009] S3. Dynamically adjust the aperture distribution according to the local clogging state of the filter screen, and at the same time adjust the working mode of the ultraviolet lamp array based on the ultraviolet transmittance and biological activity feedback;

[0010] S4. Design a linkage strategy for the backwashing cycle and the service life of the ultraviolet equipment to achieve closed-loop collaborative control of the filtration and ultraviolet links.

[0011] Preferably, in step S1, the data is obtained through a multi-dimensional sensor network. The multi-dimensional sensing network includes: a turbidity sensor and a particulate matter size detection sensor at the inlet of the filtration unit; a pressure sensor array and a biofilm monitoring unit on the surface of the filter screen; an ultraviolet transmittance probe and a flow field uniformity detection module at the inlet of the ultraviolet reactor; a microbial activity detection sensor at the outlet of the ultraviolet reactor.

[0012] Preferably, the construction of the dynamic collaborative control model includes: taking flow rate, turbidity, filter screen pressure difference gradient, ultraviolet transmittance, and biological activity concentration as input features; predicting the ultraviolet dose compensation coefficient and the filter screen aperture optimization instruction through a fuzzy neural network; and calibrating the output parameters of the model in real-time based on the biological inactivation rate at the outlet end.

[0013] Preferably, predicting the ultraviolet dose compensation coefficient and the filter screen aperture optimization instruction through a fuzzy neural network includes:

[0014] Selecting multiple key parameters as inputs, and normalizing and performing feature crossing on them;

[0015] Based on the clustering analysis of real ship data, defining fuzzy rules for several typical working conditions, and adopting a hybrid kernel function to adapt to non-linear relationships.

[0016] Preferably, dynamically adjusting the aperture distribution includes using an electrostrictive filter screen to continuously adjust the aperture according to the local pressure gradient distribution; when local blockage is detected, expanding the aperture of the blocked area and reducing the aperture of the adjacent area to maintain the overall filtration accuracy; triggering a directional gas-liquid pulse backwash, and the backwash direction is at a non-perpendicular angle to the filtration flow direction.

[0017] Preferably, adjusting the working mode of the ultraviolet lamp array includes: switching between low-frequency pulse, continuous wave, and high-frequency pulse irradiation modes according to the ultraviolet transmittance value; when the ultraviolet transmittance is lower than the threshold, activating the vortex flow generator to optimize the flow field distribution; and evenly distributing the load based on the cumulative working time of the lamp tubes.

[0018] Preferably, the linkage strategy includes: dynamically adjusting the backwash trigger threshold according to the ultraviolet lamp attenuation coefficient; predicting and shortening or extending the maintenance cycle in combination with historical water quality data; and adapting to the environmental characteristics of different sea areas through a transfer learning algorithm.

[0019] The present invention also provides a ballast water treatment-ultraviolet collaborative control system for implementing the above method, including:

[0020] An electrostrictive filter screen module configured to dynamically adjust the aperture according to the pressure gradient signal;

[0021] An ultraviolet irradiation module including a lamp array that can be controlled in zones and a vortex flow generator;

[0022] A collaborative control unit for connecting to a sensing network and performing dynamic model calculations;

[0023] A data storage module for recording the historical data of the working conditions and the adaptive learning parameters.

[0024] The beneficial effects of the present invention:

[0025] 1. Significantly improve the energy efficiency of the system

[0026] Through the dynamic collaborative control of filtration and the ultraviolet link, it avoids the energy waste caused by isolated regulation in traditional technologies. For example, when the filter is blocked, the redundant power of the ultraviolet module is synchronously reduced, and when the flow rate fluctuates, the irradiation dose is matched in real time, reducing the comprehensive energy consumption by up to 27%. The directional backwashing technology combined with the optimization of gas-liquid pulses reduces the backwashing water volume by more than 40%, improves the backwashing efficiency, and avoids the energy loss caused by ineffective backwashing.

[0027] 2. The biological inactivation rate reaches the standard stably

[0028] Based on the real-time feedback compensation mechanism of the multi-dimensional sensing network, it can quickly respond to sudden water quality changes (such as red tides and high turbidity). The dynamic adjustment delay of the ultraviolet dose is <5 seconds, and the inactivation rate is stably higher than 99.92% for a long time, far exceeding the IMOD-2 standard (inactivation rate ≥ 99.9%). The synergistic effect of gradient aperture adjustment and enhanced vortex flow field ensures double protection of the interception and inactivation of tiny plankton (10 - 40μm), avoiding the risk of missed killing in traditional single filtration or ultraviolet treatment.

[0029] 3. Extend the lifespan of key equipment

[0030] The load balancing algorithm of ultraviolet lamps makes the usage time difference of each group of lamps <50 hours, and the overall lifespan is extended to more than 9000 hours (about 6000 hours in the traditional solution). The dynamic aperture adjustment of the filter avoids local overload and blockage. Combined with the deformation energy accumulation warning mechanism, the filter replacement cycle is extended from 12 months to 18 - 24 months.

[0031] 4. Reduce the complexity and cost of operation and maintenance

[0032] The adaptive maintenance strategy triggers maintenance predictively according to the equipment status and water quality historical data, significantly reducing unplanned downtime. The transfer learning algorithm realizes cross-sea area parameter self-optimization, and no manual debugging is required when switching to a new water area (such as from the tropics to the polar regions), reducing the operation and maintenance cost.

[0033] 5. Enhance environmental adaptability

[0034] Under extreme working conditions, the system automatically adjusts control parameters through the salinity-temperature coupling correction model. The anti-biofouling design (such as capacitive biofilm monitoring) effectively inhibits the attachment of microorganisms, avoiding the risk of secondary pollution in traditional chemical cleaning. Description of the drawings

[0035] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, the drawings in the following description are only those of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0036] Figure 1 Schematic diagram of an embodiment of the present invention;

[0037] Figure 2 It is a double-layer architecture diagram of multi-modal feature fusion + prediction compensation in an embodiment of the present invention. Detailed implementation manners

[0038] To make the objectives, technical solutions and advantages of the present invention clearer and more understandable, the present invention will be further described in detail below in conjunction with specific embodiments.

[0039] It should be noted that unless otherwise defined, the technical terms or scientific terms used in the present invention should have the ordinary meanings understood by those with ordinary skills in the field to which the present invention belongs. The "first", "second" and similar terms used in the present invention do not indicate any order, quantity or importance, but are only used to distinguish different components. Words such as "including" or "comprising" mean that the elements or objects appearing before the word cover the elements or objects listed after the word and their equivalents, without excluding other elements or objects. Words such as "connected" or "coupled" are not limited to physical or mechanical connections, but may include electrical connections, whether direct or indirect. "Upper", "lower", "left", "right", etc. are only used to represent relative positional relationships, and when the absolute position of the object being described changes, the relative positional relationship may also change accordingly.

[0040] Embodiment 1:

[0041] As Figure 1 shown, an embodiment of this specification provides a ballast water treatment-ultraviolet collaborative control method and system based on multi-parameter feedback. Its core innovation lies in constructing a dynamic coupling control model, predicting and adjusting the operating parameters of the ultraviolet module by real-time analyzing the operating state of the filtration link, and simultaneously reversely optimizing the filtration strategy to form a two-way closed-loop regulation mechanism. The specific technical details include:

[0042] (1) Construction of a multi-dimensional sensing network

[0043] Arrange sensor arrays at four key nodes, namely the inlet of the filtration unit, the surface of the filter mesh, the inlet and outlet of the ultraviolet reactor, including:

[0044] Laser scattering turbidimeter (range 0-100 NTU, resolution 0.1 NTU)

[0045] High-frequency pulsed particulate matter concentration sensor (detection range 1-500 μm)

[0046] Online monitoring probe for ultraviolet transmittance (wavelength 254 nm, optical path 10 mm)

[0047] Micro pressure sensor array (arranged on the upstream side of the filter screen, with a spacing of 5 cm)

[0048] 1. Sensing layer at the inlet of the filtering unit

[0049] 1.1 Laser scattering turbidimeter

[0050] Adopting the Mie scattering compensation algorithm with dual wavelengths (650 nm / 850 nm) to eliminate the interference of seawater salinity crystallization. The sensor is equipped with a built-in temperature control cavity (accuracy of ±0.5 °C) to eliminate the influence of water temperature on the refractive index. The turbidity calculation model is as follows:

[0051]

[0052] Where I 650 , I 850 is the scattered light intensity (unit: μW / cm 2 ), which is directly measured by a photodiode. I ref is the reference light intensity (unit: μW / cm 2 ), provided by a built-in stable light source (LED wavelength 470 nm) for compensating the light source fluctuation. K1 = 2.37, K2 = 1.08 are wavelength compensation coefficients (calibrated by measured data in the Bohai Sea area) for eliminating the interference of salinity crystallization (such as NaCl) on the scattered signal. β: Temperature attenuation factor, representing the influence of water temperature on turbidity measurement, calibrated through laboratory temperature control experiments. T: Real-time water temperature (unit: °C), measured by a PT100 platinum resistance integrated in the sensor cavity, with an accuracy of ±0.1 °C.

[0053] 1.2 High-frequency pulsed particulate matter concentration sensor

[0054] Based on the principle of acoustic resonance, a pulse sequence of 20 - 200 kHz is emitted, and the particle size distribution is analyzed through the echo frequency shift characteristics:

[0055]

[0056] In the formula, Δf is the echo signal frequency shift amount (unit: Hz), calculated from the difference between the emission frequency f0 (reference value 200 kHz) and the received frequency. v is the fluid flow velocity (unit: m / s), inverted through the Doppler frequency shift formula , where c is the sound wave propagation speed in seawater (unit: m / s), and the calculation formula is c = 1449 + 4.6T - 0.055T 2 + 1.34(S - 35) (T is the water temperature, S is the salinity). A i : The amplitude weight coefficient of the i-th particle size interval (10 - 15 μm, 15 - 40 μm, 40 - 80 μm), extracted by wavelet packet decomposition (WPD) to obtain the energy proportion E 10-15 , E15-40 , E 40-80 (Unit: %). C cal : The 3×3 calibration coefficient matrix is obtained through laboratory calibration of sediment samples from the Yangtze Estuary (median particle size D50 = 32 μm) and is used to convert the energy proportion into the actual mass concentration (unit: mg / L).

[0057] Wavelet packet decomposition is used to extract the energy proportion of each frequency band, and a particle size distribution matrix is established:

[0058]

[0059] In the formula, fecho represents the original frequency spectrum of the received echo signal (unit: Hz), C cal is the calibration coefficient matrix of the sediment in the Yangtze Estuary.

[0060] 2. Filter surface condition perception layer

[0061] 2.1 Micro pressure sensor array

[0062] An MEMS piezoresistive sensor (range 0 - 50 kPa, response time < 1 ms) is embedded in the ZrO2 - TiN filter substrate. The local blockage coefficient is calculated through the spatial pressure gradient:

[0063]

[0064] In the formula, P i，j is the real - time pressure value measured by the pressure sensor at the i - th row and j - th column on the filter surface (unit: kPa). Δx is the grid spacing of the sensor array (unit: cm), which is optimally fixed at 5 cm for the discretized calculation of the spatial derivative. is the local pressure gradient (unit: kPa / cm 2 ), which reflects the spatial non - uniformity of the filter blockage. When , it is determined as mechanical blockage caused by hard particulate matter (such as shellfish larvae), and the aperture dynamic adjustment is triggered.

[0065] 2.2 Capacitive biofilm monitoring unit

[0066] Interdigitated electrodes (spacing 200 μm) are arranged on the backflow surface of the filter, and microbial attachment is detected through dielectric spectroscopy (frequency sweep 1 kHz - 10 MHz):

[0067]

[0068] Where ω is the angular frequency (unit: rad / s), generated by a frequency sweep signal source, with a range of 1 kHz - 10 MHz. C0 is the initial capacitance (unit: F), calibrated in the state of a clean filter screen, and the typical value is 120 pF. Z is the circuit impedance (unit: Ω), measured by a lock-in amplifier. V in , V out and V bio bio are the input and output voltages (unit: V), used to calculate the signal attenuation. ε″ 254，sampl

[0069] 3. UV Reactor Inlet Sensing Node

[0070] 3.1 Online Monitoring of Ultraviolet Transmittance (UVT)

[0071] Adopt a dual - optical - path differential measurement structure (254 nm / 365 nm) to eliminate the interference of organic pigments:

[0072]

[0073] Where I 254，sampl e is the transmitted light intensity of the 254 - nm ultraviolet light in the water sample to be measured (unit: μW / cm 2 ). I 365，ref is the transmitted light intensity of the 365 - nm reference light in a standard quartz cuvette (unit: μW / cm 2 ), used to correct the light source fluctuation. I 254，ref is the transmitted light intensity of the 254 - nm ultraviolet light in a standard glycerol solution (UVT = 95%), used as a reference value.

[0074] Optical path adaptive adjustment algorithm:

[0075]

[0076] Dynamically adjust the probe spacing through a stepper motor to ensure that the signal - to - noise ratio > 40 dB.

[0077] 3.2 Flow Field Distortion Monitoring Module

[0078] Arrange a 4×4 ultrasonic Doppler velocimetry array to construct a flow field uniformity index:

[0079]

[0080] Where v i , j is the real - time flow velocity (unit: m / s) of the velocity measurement point in the i - th row and j - th column of the UV reactor, measured by an ultrasonic Doppler probe. is the average flow velocity at the reactor cross-section (unit: m / s), calculated as the arithmetic mean of 16 velocity measurement points. U index is the flow field uniformity index (dimensionless). When U index <0.85, the vortex generator is activated, and the deflection angle θ = 15°·(1 - U index ).

[0081] 4. UV Reactor Outlet Verification Layer

[0082] 4.1 On-line Microbial Activity Detector

[0083] Adopt the fluorescence excitation method (excitation wavelength 405nm, detect the chlorophyll a signal at 680nm), combined with the pulse amplitude modulation (PAM) technology:

[0084]

[0085] Where F0 is the initial fluorescence value (unit: relative fluorescence unit, RFU), measured under the state of no photochemical quenching. F m is the maximum fluorescence value (unit: RFU), measured under the excitation of a saturating pulse light (3000 μmol photons / m 2 / s). F v / F m is the maximum photochemical efficiency of photosystem II (dimensionless). When F v / F m <0.3, it is determined that the algal cell membrane is ruptured and the inactivation is up to standard.

[0086] 4.2 Residual Oxidant Monitoring

[0087] For the UV / hydrogen peroxide combined system, deploy a platinum microelectrode array to detect the residual H2O2:

[0088] i lim =4nFDc·(1 + 0.23Re 0.65 Sc 0.33 ),

[0089] Where n is the number of electron transfers (n = 2 in the H2O2 oxidation reaction). F is the Faraday constant (96485 C / mol). D is the diffusion coefficient (unit: m 2 / s), related to the water temperature (at 25°C, D of H2O2 = 1.4×10 -9 m 2 / s). Re is the Reynolds number ( is the density, d is the electrode diameter). Sc is the Schmidt number By measuring i lim (unit: A), the H2O2 concentration c (unit: ppm) is back-calculated to ensure c < 0.1 ppm.

[0090] 5. Data fusion architecture

[0091] Adopt the spatio-temporal alignment algorithm to unify the data of each node to the fluid element tracking coordinate system:

[0092]

[0093] where is the spatial coordinate of the fluid element at time t (unit: m). is the initial position of the element (unit: m), recorded by the inlet sensor. is the flow velocity vector of the element at time T (unit: m / s), obtained by interpolating the Doppler velocity measurement array.

[0094] Establish the material transport correlation from the filter screen to the ultraviolet through the particle tracking model (PTM) to realize the causal reasoning across links. In the actual measurement of Zhoushan Port, the detection rate of copepod larvae of 40 - 60 μm reaches 99.2%, and the turbidity monitoring error < ±2.5%. Compared with the traditional single-point sensing scheme, the data dimension is increased by 4 times, providing accurate input for collaborative control.

[0095] (2) Dynamic collaborative control model

[0096] Aiming at the core challenges of strong water quality time-variability (such as salinity fluctuations and sudden changes in suspended solid concentration) and equipment dynamic coupling (nonlinear correlation between filter screen blockage and ultraviolet dose requirement) in ship ballast water treatment, this model adopts a two-layer architecture of multi-modal feature fusion + prediction compensation, as Figure 2 shown, and realizes the collaborative optimization across links through a hybrid strategy of physical mechanism modeling and data-driven learning.

[0097] 1. Construction of composite state equation

[0098] The model takes the ultraviolet dose requirement as the core output variable and establishes a three-dimensional coupling equation including fluid mechanics, optical transmission, and biological inactivation:

[0099]

[0100] -Q: Real-time flow rate (m3 / h), measured by an electromagnetic flowmeter.

[0101] -η UV : Ultraviolet lamp efficiency decay coefficient (0.8 - 1.0, related to the cumulative working time t UV related:

[0102]

[0103] -A eff : Effective irradiation area of the ultraviolet reactor (m 2) Calculated dynamically according to the number of activated lamp arrays.

[0104] -ΔP: Real-time differential pressure of the filter (kPa), ΔP max is the maximum tolerable differential pressure of the filter (50 kPa).

[0105] -C bio : Bioactive concentration (μg / L), retrieved by inverting the chlorophyll a signal detected by the fluorescence sensor at the outlet end.

[0106] -K bio = 12.5 μg / L: Half-saturation constant, calibrated through the inactivation experiment of common algae (Skeletonema costatum) in the Bohai Sea.

[0107] 2. Implementation of the improved fuzzy neural network

[0108] 2.1 Feature engineering of the input layer

[0109] Select 8 key parameters as inputs and perform normalization and feature crossing on them:

[0110] (Flow normalization)

[0111] x2 = ln(T + 1) (Logarithmic transformation of water temperature)

[0112] (Proportion of small particles)

[0113] (Maximum local differential pressure gradient)

[0114] (Transmittance normalization)

[0115] (Bioactive saturation function)

[0116] 2.2 Generation of fuzzy rules

[0117] Based on the clustering analysis of real ship data, define the fuzzy rules for 5 typical working conditions:

[0118] - Rule 1: If x3 > 0.6 and x4 > 0.8, then trigger the gradient filtration mode:

[0119] Aperture adjustment = 10 + 15·tanh(2x3) + 5x4 (μm)

[0120] At the same time, the ultraviolet dose compensation coefficient increases by 30%.

[0121] - Rule 2: If x5 < 0.4 and x7 < 0.6, then activate the vortex strengthening mode:

[0122] Vortex rotation speed = 500 + 300·(0.6 - x7) (rpm)

[0123] The ultraviolet lamp tube is switched to the high - frequency pulse mode (duty cycle of 200Hz).

[0124] 2.3 Radial basis function (RBF) hidden layer

[0125] Adopt a hybrid kernel function to adapt to the non - linear relationship:

[0126]

[0127] -c j : Clustering center, divided from historical operating condition data by K - means.

[0128] -σ j : Kernel width, adaptively adjusted according to the within - class sample variance.

[0129] -γ = 0.3: Weight of the Sigmoid term, enhancing the response ability to mutation operating conditions.

[0130] 3. Dynamic compensation mechanism

[0131] 3.1 Ultraviolet dose feedback calibration

[0132] Collect the microbial inactivation rate η at the outlet end every 5 seconds kill , if η kill <99.8%, start PID compensation:

[0133]

[0134] The coefficient is dynamically adjusted according to UVT:

[0135] K p =0·5·UVT / 100, K i =0.02·e -0.1(100-UVT) ,K d =0.1·(100 - UVT).

[0136] 3.2 Adaptive adjustment of filter screen aperture

[0137] Realize local precise regulation based on the pressure gradient distribution:

[0138] - Hard particle blockage (such as shellfish larvae, etc.):

[0139]

[0140] The aperture of adjacent areas is reduced to maintain the overall filtration accuracy.

[0141] - Soft biofilm attachment: When ε″ is detected bioWhen it is > 0.02, trigger local electrolytic cleaning and temporarily expand the pore size by 10 μm to maintain the flux.

[0142] 4. Sea area adaptability optimization

[0143] Preset control strategies according to different water area characteristics:

[0144] 1) High sediment estuary areas (such as the Yangtze River Estuary):

[0145] - Activate the turbidity feed-forward channel. When the turbidity > 50 NTU, the UV dose is pre-increased by 20%.

[0146] - The filter screen backwashing cycle is shortened to:

[0147]

[0148] 2) Low-temperature and high-salinity sea areas (such as the Arctic shipping route):

[0149] Correct the salinity influence term in the calculation of UV transmittance:

[0150] UVT adj = UVT·(1 - 0.0035·S) (S: salinity, psu) $

[0151] - The lower limit of the filter screen pore size is adjusted to 15 μm to prevent the penetration of sea ice microcrystals.

[0152] Through the deep integration of mechanism and data, this model realizes the intelligent control paradigm from "passive response" to "prediction - coordination".

[0153] (3) Adaptive adjustment mechanism

[0154] 1. Adaptive adjustment in the filtration link

[0155] 1.1 Electrostrictive filter screen dynamic pore size regulation

[0156] Adopt ZrO2 - TiN composite piezoelectric material. By applying an alternating electric field (0 - 200 V), change the lattice structure to achieve continuous pore size adjustment (10 - 80 μm). The relationship between the strain Δd and the electric field strength E is:

[0157]

[0158] -k p : Piezoelectric coefficient (optimized by doping 3% Y2O3)

[0159] -T c : Curie temperature of the material to ensure stability at the ship's environmental temperature (-20 ~ 50 °C)

[0160] -d0: Initial pore size (μm)

[0161] Dynamic adjustment strategy:

[0162] 1) Local blockage response: When the pressure sensor detects the pressure difference gradient in a certain area :

[0163] Instantaneously enlarge the aperture of this area to:

[0164]

[0165] Aperture reduction compensation in adjacent areas:

[0166]

[0167] 2) Gradient filtration mode: Under high turbidity conditions (SS>100mg / L), generate an aperture distribution that is sparse in the front and dense in the back:

[0168] d(x) = 10 + 70·(1 - e -0.1x (x is the normalized distance along the water flow direction)

[0169] 1.2 Optimization of directional backwashing

[0170] Trigger condition: Combine the blockage index with the biofilm thickness δ bio (monitored by capacitance):

[0171] Backwashing trigger threshold: CI > 25 or δ bio > 50μm

[0172] Pulse parameter design:

[0173] - Gas-liquid mixing ratio:

[0174] φ = 0.6 + 0.1·ln(1 + SS / 10) (gas volume ratio)

[0175] Dynamic calculation of a pulse pressure:

[0176]

[0177] The flushing direction forms a 45° angle with the water flow, and the shear force is used to improve the stripping efficiency. Experiments show that the efficiency is increased by 38% compared with vertical flushing.

[0178] 2. Dynamic compensation in the UV link

[0179] 2.1 Nonlinear compensation model for UV dose

[0180] Basic dose calculation: Based on the improved optical hydrodynamics equation:

[0181]

[0182] -P UV : Lamp radiant power (W / cm 2 ), which decays with lifetime:

[0183] P UV (t) = P0·e -0.00015t (t is working time, hours)

[0184] -η opt : Optical system efficiency (0.65-0.85), related to the cleanliness of the quartz sleeve.

[0185] Dynamic compensation strategy:

[0186] 1) Real-time UVT compensation: When UVT decreases, the dose is increased exponentially:

[0187]

[0188] Simultaneously activate the vortex generator, speed:

[0189] ω=300+500·(1-UVT / 100)(rpm)

[0190] 2) Biological activity feedback compensation: according to the outlet F v / F m Value to fine tune dosage:

[0191]

[0192] 2.2 Intelligent Scheduling of Lamp Array

[0193] Group start-stop strategy: UV lamps are divided into 4 groups (G1-G4) and dynamically switched based on real-time demand:

[0194] UVT range Activation group Working mode ≥90％ G1 Low-frequency pulse (50Hz) 80％-90％ G1+G2 Alternating pulse (100Hz) 70％-80％ G 1+G2+G3 Continuous wave <70％ Full group High-frequency pulse (200Hz)

[0195] -Life balance algorithm: record the cumulative working time t of each group Gi , enable t first Gi The smallest group, the standard deviation is controlled at σ t <50 hours.

[0196] 3. Collaborative maintenance strategy

[0197] 3.1 Backwash cycle linkage optimization

[0198] Establish a correlation model between filter clogging rate and UV lamp decay:

[0199]

[0200] -K_1=120,K_2=0.03: Calibrated by data from Bohai Sea and South China Sea

[0201] -SS avg : Average suspended solid concentration in the past 24 hours (mg / L)

[0202] 3.2 Preventive maintenance trigger mechanism

[0203] - Ultraviolet lamp replacement warning: When the lamp efficiency η UV <0.7 or the current fluctuation ΔI>10%, automatically switch to the standby lamp group.

[0204] - Filter life prediction: Based on the cumulative deformation energy model:

[0205]

[0206] When E total >1.5×10 4 J / m 2 prompt to replace the filter.

[0207] 4. Special application scenario adaptation

[0208] 4.1 High-sediment estuary (such as the Yangtze Estuary)

[0209] - Aperture adjustment compensation: When SS>150mg / L, start the aperture fluctuation mode:

[0210] d(t)=d base +5·sin(2πft)(f=0.2Hz)

[0211] Prevent sediment caking by periodically fine-tuning the aperture.

[0212] - Ultraviolet dose pre-boost: According to turbidity feedforward:

[0213] Dose pre =Dose base ·(1 + 0.015·NTU)

[0214] 4.2 Low-temperature and high-salinity sea area (such as the Arctic shipping route)

[0215] - Material deformation compensation: Correct the temperature term in the piezoelectric coefficient:

[0216]

[0217] - Salinity-UVT coupling correction:

[0218] UVT real =UVT measured ·(1 - 0.004S)(S: salinity, psu)

[0219] This mechanism realizes efficient collaborative control under complex working conditions through multi-physical field coupling modeling and online learning optimization.

[0220] Example Two:

[0221] The specific implementation process of the above system includes:

[0222] Step 1: System initialization and dynamic calibration

[0223] Hardware self-check and reset

[0224] After the system is powered on, the control unit automatically performs hardware diagnosis:

[0225] The electrostrictive filter is reset to the minimum pore size (10 μm), and the ultraviolet lamp array enters the preheating state (70% of the reference power).

[0226] The micro pressure sensor array scans the surface of the filter, records the initial pressure distribution matrix P0P0, and establishes the reference flow resistance curve.

[0227] The ultraviolet transmittance probe performs dual-path calibration: zero-point and full-scale calibration are carried out using a standard glycerol solution (UVT = 95%) and air (UVT = 100%) to eliminate the optical window pollution error.

[0228] Environmental parameter adaptation

[0229] Through the salinity sensor (electrode method) and the water temperature probe, the preset sea area parameter library (such as the Yangtze River Estuary, Arctic shipping lane) is loaded, and the initial control mode is automatically matched.

[0230] If the detected salinity > 30 psu and the water temperature < 5 °C, the "high salinity and low temperature mode" is activated, and the salinity compensation coefficient in the ultraviolet dose calculation is adjusted to 1.2 times.

[0231] Step 2: Multi-source data fusion and feature extraction

[0232] Real-time data acquisition and noise reduction

[0233] Eight types of parameters such as the pressure difference gradient of the filter, ultraviolet transmittance, and biological activity concentration are synchronously collected every 200 ms, and impulse interference is eliminated through Kalman filtering.

[0234] The signal of the particulate matter sensor is decomposed by wavelet packet, and the Doppler frequency shift component (for flow velocity calculation) and the particle size characteristic frequency bands (10 - 15 μm, 15 - 40 μm, 40 - 80 μm) are separated.

[0235] Quantification of working conditions

[0236] Calculate the filter clogging index: based on the weighted sum of the pressure gradient matrix and the suspended solid concentration, when the index exceeds the threshold of 1.2, it is determined to enter the "high clogging risk state".

[0237] Construct the ultraviolet dose requirement matrix: Divide the ultraviolet reactor into 8 control zones along the fluid path, and generate a differential dose compensation coefficient according to the UVT value and the flow field uniformity index of each zone.

[0238] Step 3: Dynamically execute the coordinated control instructions

[0239] Dynamic adjustment of the filtration link

[0240] Local aperture adjustment: When the pressure difference gradient in a certain area suddenly increases by 20%, the control unit applies a 200V pulse voltage to this area, instantaneously expanding the aperture to 15 - 25μm, while reducing the aperture of the adjacent area by 2 - 5μm, forming a "sparse - dense alternating" filtration gradient to balance flux and precision.

[0241] Directional backwash trigger: Use a gas - liquid mixed pulse (nitrogen accounts for 60%, pressure 0.5 - 0.8MPa), the flushing direction forms a 45° angle with the filter screen plane, and preferentially removes hard particles such as shellfish larvae. The backwash duration is dynamically adjusted according to the blockage index (50 - 80ms) to avoid excessive backwashing and loss of the filter screen.

[0242] Intelligent regulation of the ultraviolet link

[0243] Lamp array combination strategy: When UVT < 70%, fully start the ultraviolet lamp group and activate the vortex flow generator (rotation speed 500 - 800rpm), forcing the water flow to form a spiral path to extend the irradiation time.

[0244] Dose closed - loop calibration: Read the data of the bio - activity sensor at the outlet end every 5 seconds. If the chlorophyll a fluorescence signal Fv / Fm > 0.4, increase the dose in 10% steps until the inactivation rate is stable above 99.8%. At the same time, record the overshoot for optimizing the control parameters of the next similar working condition.

[0245] Step 4: Energy efficiency optimization and cross - link maintenance

[0246] Energy consumption balance management

[0247] Generate the energy consumption - efficiency ratio (EER) curve daily, and automatically optimize the start - stop sequence of ultraviolet lamps and the adjustment range of the filter screen aperture. For example, during the low - load period at night, preferentially enable the lamp group with less cumulative working time to extend the overall life.

[0248] When it is detected that the current fluctuation of the ultraviolet module exceeds ±8%, automatically switch to the standby lamp group and mark the abnormal lamp as the to - be - maintained state.

[0249] Preventive maintenance linkage

[0250] According to the correlation model between the cumulative deformation energy of the filter screen and the attenuation coefficient of the ultraviolet lamp tube, predictively trigger the maintenance cycle:

[0251] If the deformation energy of the filter screen > 10000 J / m 2 And the ultraviolet lamp efficiency < 75%, it is recommended to replace the filter screen and the aging lamp tube synchronously.

[0252] Before entering high turbidity sea areas (such as the Yangtze River Estuary), automatically shorten the backwashing cycle to 60% of the normal value, and pre-increase the ultraviolet dose by 10%.

[0253] Adaptive learning across sea areas

[0254] When the ship enters a new sea area, the system migrates the most similar operating condition data (such as salinity, turbidity, biological species) in the historical database and completes the self-optimization of control parameters within 2 hours. For example, when switching from the East China Sea to the South China Sea waters, automatically reduce the initial aperture of the filter screen to 12 μm to cope with more microplankton.

[0255] Step Five: Fault Diagnosis and Emergency Handling

[0256] Multi-level alarm mechanism

[0257] Primary alarm: When the key parameters (such as UVT, inactivation rate) deviate from the set range by 10%, trigger an audible and visual alarm and automatically switch to the redundant control module.

[0258] Secondary emergency: If the ultraviolet module fails completely, immediately start the high-pressure filtration mode (aperture reduced to 8 μm), and at the same time send a maintenance request to the shipboard central control system.

[0259] Data traceability and root cause analysis

[0260] Store the full-dimensional operating condition data of the past 72 hours. When an anomaly occurs, automatically generate a fault tree analysis report. For example, if backwashing is frequently triggered, the system will conduct a correlation analysis on the sudden increase event of the suspended solid concentration and the spatio-temporal distribution characteristics of the pressure difference gradient, and prompt local structural defects of the filter screen or sensor drift.

[0261] Implementation effect

[0262] In the in-ship verification (Qingdao-Rotterdam route in 2023), this solution demonstrates the following advantages:

[0263] Response agility: The full-process response time from detecting filter screen blockage to completing ultraviolet dose compensation < 5 seconds, which is 6 times faster than the traditional system.

[0264] Collaborative optimization ability: Under the high turbidity operating conditions in the Yangtze River Estuary, the system reduces the energy consumption of the ultraviolet module by 35% through dynamic aperture adjustment, while maintaining an inactivation rate of 99.91%.

[0265] Intelligent maintenance: The predictive maintenance function reduces the unplanned downtime by 82%, and the filter screen replacement cycle is extended to 18 months (the traditional solution is 12 months).

[0266] Through the deep cooperation mechanism, this embodiment realizes the leapfrog upgrade of the ballast water treatment system from "passive execution" to "active optimization".

[0267] Those of ordinary skill in the art should understand that the discussion of any above embodiment is only exemplary and is not intended to imply that the scope of the present invention is limited to these examples; under the concept of the present invention, the technical features in the above embodiments or different embodiments can also be combined, the steps can be implemented in any order, and there are many other variations of different aspects of the present invention as described above, and they are not provided in detail for the sake of brevity. Any omission, modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included within the protection scope of the present invention.

Claims

1. A ballast water treatment-ultraviolet collaborative control method based on multi-parameter feedback, characterized in that The method includes the following steps: S1. Obtain water quality parameters, filter screen working conditions, and ultraviolet irradiation efficiency data; S2. Based on the data obtained in step S1, construct a dynamic collaborative control model, predict the ultraviolet dose requirement through a fuzzy neural network, and generate a filter screen aperture adjustment instruction; S3. Dynamically adjust the aperture distribution according to the local clogging state of the filter screen, and at the same time adjust the working mode of the ultraviolet lamp array based on the feedback of ultraviolet transmittance and biological activity; S4. Design a linkage strategy for the backwashing cycle and the service life of the ultraviolet equipment to achieve closed-loop collaborative control of the filtration and ultraviolet links.

2. The ballast water treatment-ultraviolet collaborative control method based on multi-parameter feedback according to claim 1, wherein, In step S1, the data is obtained through a multi-dimensional sensor network. The multi-dimensional sensor network includes: a turbidity sensor and a particulate matter size detection sensor at the inlet of the filtration unit; a pressure sensor array and a biofilm monitoring unit on the surface of the filter screen; an ultraviolet transmittance probe and a flow field uniformity detection module at the inlet of the ultraviolet reactor; a microbial activity detection sensor at the outlet of the ultraviolet reactor.

3. The ballast water treatment-ultraviolet collaborative control method based on multi-parameter feedback according to claim 1, wherein The construction of the dynamic collaborative control model includes: taking flow rate, turbidity, filter screen pressure difference gradient, ultraviolet transmittance, and biological activity concentration as input features; predicting the ultraviolet dose compensation coefficient and the filter screen aperture optimization instruction through a fuzzy neural network; and calibrating the output parameters of the model in real-time based on the biological inactivation rate at the outlet end.

4. The ballast water treatment-ultraviolet collaborative control method based on multi-parameter feedback according to claim 3, wherein, The prediction of the ultraviolet dose compensation coefficient and the filter screen aperture optimization instruction through the fuzzy neural network includes: Selecting multiple key parameters as inputs, and normalizing and performing feature crossing on them; Based on the clustering analysis of actual ship data, defining fuzzy rules for several typical working conditions, and using a hybrid kernel function to adapt to the non-linear relationship.

5. The ballast water treatment-ultraviolet collaborative control method based on multi-parameter feedback according to claim 1, characterized in that The dynamic adjustment of the aperture distribution includes using an electrostrictive filter screen to continuously adjust the aperture according to the local pressure gradient distribution; when local clogging is detected, expanding the aperture in the clogging area and narrowing the aperture in the adjacent area to maintain the overall filtration accuracy; triggering a directional gas-liquid pulse backwashing, and the backwashing direction is at a non-perpendicular angle to the filtration flow direction.

6. The ballast water treatment-ultraviolet collaborative control method based on multi-parameter feedback according to claim 1, characterized in that The adjustment of the working mode of the ultraviolet lamp array includes: switching the low-frequency pulse, continuous wave, and high-frequency pulse irradiation modes according to the ultraviolet transmittance value; when the ultraviolet transmittance is lower than the threshold, activating the vortex flow generator to optimize the flow field distribution; and evenly distributing the load based on the cumulative working time of the lamps.

7. The ballast water treatment-ultraviolet collaborative control method based on multi-parameter feedback according to claim 1, characterized in that The linkage strategy includes: dynamically adjusting the backwashing trigger threshold according to the ultraviolet lamp attenuation coefficient; predicting and shortening or extending the maintenance cycle in combination with historical water quality data; and adapting to the environmental characteristics of different sea areas through a transfer learning algorithm.

8. A ballast water treatment-ultraviolet collaborative control system for implementing the method according to any one of claims 1-7, characterized in that, It includes: An electrostrictive filter screen module configured to dynamically adjust the aperture according to the pressure gradient signal; An ultraviolet irradiation module including a lamp array and a vortex flow generator that can be controlled in zones; A collaborative control unit for connecting the sensor network and performing dynamic model calculations; A data storage module for recording the historical data of the working conditions and the adaptive learning parameters.

Images