Submerged plant intelligent light supplementing system based on optical fiber and control method thereof

Abstract

The invention discloses a submerged plant intelligent light supplementing system based on an optical fiber and a control method thereof, and the method comprises the steps: obtaining an environment image and environment data of a submerged plant, segmenting the environment image to obtain a regional environment image, carrying out the image recognition of the regional environment image, calculating the biological richness according to an image recognition result, and determining the regional illumination demand. The method comprises the steps of performing feature extraction on a growth environment, determining a first light supplementing demand and a second light supplementing demand, constructing an intelligent submerged plant light supplementing model, inputting the first light supplementing demand and the second light supplementing demand into the intelligent submerged plant light supplementing model to obtain a light supplementing strategy, and performing intelligent light supplementing on submerged plants according to the light supplementing strategy. The method not only can improve the precision and efficiency of the intelligent light supplement of the submerged plant by the optical fiber, but also has good interpretability, and can be directly applied to the intelligent light supplement system of the submerged plant by the optical fiber.

Description

Technical Field

[0001] The present invention relates to the field of intelligent lighting supplement, and in particular to an optical fiber-based submerged plant intelligent lighting supplement system and a control method thereof. Background Art

[0002] Submerged plants absorb nutrients from the water to alleviate eutrophication during their growth, and release oxygen through photosynthesis to increase the dissolved oxygen content in the water. They also have certain economic value and play an important role in ecological restoration, aquatic plant cultivation, and aquarium viewing. The optical fiber light supplement system has good light-guiding performance and durability. Using optical fiber as a transmission medium, the light signal emitted by the light source can be transmitted through optical fiber to areas that are far away or difficult to directly illuminate and need light supplement.

[0003] Since natural light decays rapidly underwater and emergent plants block natural light, submerged plants often grow slowly or even die due to insufficient light, and submerged plants need to be supplemented with light. Traditional submerged plant supplementary lighting equipment mostly relies on manual placement and salvage, requires frequent maintenance and replacement, and is difficult to adjust the light intensity, spectral range and irradiation area in a timely and flexible manner. By fully considering the growth environment of submerged plants, obtaining the supplementary lighting needs of submerged plants, and combining optical fiber technology, a timely, accurate and efficient optical fiber-based submerged plant intelligent supplementary lighting system and its control method are designed to overcome the shortcomings of existing supplementary lighting systems and control technologies, and to accurately and efficiently supplement the lighting of submerged plants, which is of great significance to maintaining the balance of the ecosystem and creating economic value. Summary of the invention

[0004] The object of the present invention is to provide an optical fiber-based submerged plant intelligent lighting system and a control method thereof.

[0005] To achieve the above object, the present invention is implemented according to the following technical solutions:

[0006] The present invention comprises the following steps:

[0007] Acquire environmental images and environmental data of submerged plants, and pre-process the environmental images and environmental data; the environmental data includes growth environment and lighting environment;

[0008] The environmental image is segmented according to the submerged plant growth area to obtain a regional environmental image, image recognition is performed on the regional environmental image to obtain an image recognition result, and the biological richness is calculated according to the image recognition result; the image recognition result includes the number of plants, plant types, plant growth stages and water quality environment;

[0009] Matching the plant category and the plant growth stage with a standard plant information database to obtain the lighting conditions of submerged plants, and determining the regional lighting requirements according to the lighting conditions, the number of plants and the lighting environment data;

[0010] Extracting features of the growth environment to obtain underwater temperature, oxygen content, and carbon dioxide content, determining a first supplemental lighting requirement according to the underwater temperature, the biological richness, and the regional lighting requirement, and determining a second supplemental lighting requirement according to the oxygen content, the carbon dioxide content, the water quality environment, and the regional lighting requirement;

[0011] A submerged plant intelligent fill-light model is constructed, the first fill-light requirement and the second fill-light requirement are input into the submerged plant intelligent fill-light model to obtain a fill-light strategy, and intelligent fill-light is performed on the submerged plants according to the fill-light strategy.

[0012] Furthermore, the method of segmenting the environmental image according to the submerged plant growth area to obtain a regional environmental image includes:

[0013] The growth range of the submerged plants is projected onto a two-dimensional plane, the cell area is determined according to the fill light range of the optical fiber fill light device, and the growth area of ​​the submerged plants is divided according to the cell area;

[0014] The environmental image is segmented according to the growth area of ​​the submerged plants to obtain a regional environmental image, and regional coordinates are set for the regional environmental image.

[0015] Furthermore, the method for performing image recognition on the regional environment image includes:

[0016] (1) Construct a support vector machine image classification and counting model, input the regional environmental image into the support vector machine image classification and counting model, output the classification and counting results of plants and animals in the regional environmental image, and record the types and numbers of submerged plants;

[0017] (2) The phenological extraction method of time series NDVI is used to obtain the growth stages of submerged plants, including:

[0018] Draw a region of interest enclosed by 3 or more points on the regional environment image, and calculate the average values ​​of the three color bands in the region of interest;

[0019] Extract vegetation indices that characterize the growth characteristics of submerged plants, including the normalized difference vegetation index NDVI, relative greenness index GI, relative blueness index BI, relative redness index RI, and relative blue-red brightness BRI index;

[0020] The inverse Fourier transform function ifft was used to reconstruct the time series NDVI, and the NDVI points within the interval were fitted to obtain the seasonal trajectory curve of submerged plant growth. The root mean square error was used to evaluate the fitting effect.

[0021] The growth season trajectory curve of submerged plants is expressed as:

[0022]

[0023] Where c1 is the mean value of light and vegetation fraction in winter, c2 is the difference between light and vegetation fraction in summer and winter, c3 is the trend of greenness in summer, t is time, x1 is the inflection point of greenness increase in spring, x3 is the inflection point of greenness decrease in autumn, x2 and x4 are parameters for adjusting the seasonal growth curves in spring and autumn, x2 and x4∈(0,3];

[0024] According to the seasonal trajectory curve of submerged plant growth, phenological parameters are extracted to obtain plant growth stages and corresponding growth parameters;

[0025] (3) Identify impurities in the water in the regional environmental image to determine the water quality environment, perform Gaussian filtering to denoise the regional environmental image, and convert it into a grayscale image. Calculate the three Tamura texture features of the grayscale image, namely, medium roughness, directionality, and uniformity. Perform a comprehensive evaluation based on the set roughness threshold, directionality threshold, and uniformity threshold to obtain a water quality score. Combine the water quality score, pH value, and salt ion concentration to determine the water quality environment.

[0026] Furthermore, the method for calculating biological richness according to the image recognition result includes:

[0027] Substitute the species category and species number obtained by image recognition into the biological richness function to obtain the biological richness. The expression of the biological richness function is:

[0028]

[0029] Where B is the biological richness, T is the number of all organisms in the area, C is the total number of all biological categories in the area, A is the area, and n is the total area. i 、n j 、n k are the number of aquatic plants of the ith category, aquatic animals of the jth category and microorganisms of the kth category, w1, w2, w3 are the category weights of aquatic plants, aquatic animals and microorganisms, l, m, n are the number of categories of aquatic plants, aquatic animals and microorganisms, respectively.

[0030] Furthermore, the method for determining the regional lighting requirement includes:

[0031] According to the plant category and plant growth stage, the plant category and plant growth stage of the standard plant information database are matched, and the similarity between the submerged plants to be matched and the submerged plants in the standard plant information database is calculated one by one, and the light conditions of the submerged plants in the standard plant information database corresponding to the matching group with the highest similarity are extracted;

[0032] Lighting conditions include ideal spectral range and ideal light intensity. Feature extraction of lighting environment data is performed to obtain light intensity, spectral range, photon flux density, and radiation flux density. The ideal light intensity is adjusted according to the lighting environment data and the number of plants, and the regional light demand is calculated. The expression of regional light demand is:

[0033]

[0034] Among them I y ′ is the required light intensity in the area, α and β are the light parameter influencing factors, PPFD is the light quantum flux density, N is the number of all submerged plants in the area, l is the number of submerged plant categories, I yi is the ideal light intensity for the i-th type of submerged plants, PAR is the radiation flux density, w 1i is the weight coefficient of the i-th type of submerged plants, n i is the number of submerged plants of type i;

[0035] The spectral ranges of all submerged plants are sorted according to the size of the values, the largest starting wavelength is taken as the spectrum end judgment point, and the smallest ending wavelength is taken as the spectrum starting judgment point. The regional required spectral range is determined based on the spectrum end judgment point and the spectrum starting judgment point, and the regional required light intensity and the regional required spectral range are combined to determine the regional light demand.

[0036] Furthermore, the method for determining the first fill light requirement includes:

[0037] The environmental factors that affect the photosynthesis of submerged plants are determined based on the underwater temperature and biological richness. The underwater temperature, the ideal temperature of submerged plants and biological richness are substituted into the environmental function to obtain the environmental factors. The environmental function expression is:

[0038]

[0039] Where E is the environmental factor, x1 and x3 are the positive coefficients of the influence of the adjustment factor, T is the underwater temperature, and T y is the ideal temperature corresponding to the maximum efficiency of photosynthesis of submerged plants, x2 is the sensitivity of temperature change to environmental factors, and there is no linear relationship between environmental factors and the first term, B is biological richness, x4 is the sensitivity of biological richness to environmental factors, and the second term is the attenuation coefficient, and environmental factors are negatively correlated with it;

[0040] The first fill light requirement is determined according to environmental factors and regional lighting requirements. The corresponding expression of the lighting intensity of the first fill light requirement is:

[0041]

[0042] Where ΔI1 is the light intensity required for the first fill light, I y ′ is the required light intensity of the area, I t is the light intensity corresponding to the underwater lighting environment;

[0043] The spectral range required by the first fill light is consistent with the spectral range required by the area.

[0044] Furthermore, the method for determining the second fill light requirement includes:

[0045] Substituting the oxygen content, carbon dioxide content and water quality environment in the water into the second fill light demand function, the light intensity required for the second fill light is obtained. The expression of the second fill light demand function is:

[0046]

[0047] Among them, ΔI2 is the light intensity required for the second supplementary light, k1 and k2 are the influence coefficients of oxygen and carbon dioxide, d1 and d2 are the oxygen content and carbon dioxide content in water, r1 and r2 are the response coefficients of oxygen and carbon dioxide, p and q are the changes in oxygen content and carbon dioxide content, PH and PH y is the pH value in water and the ideal pH value, c, c y is the nutrient ion concentration and ideal nutrient ion concentration in water, o, o y to rate water quality and ideal water quality;

[0048] The spectral range required by the second fill light is consistent with the spectral range required by the area.

[0049] Furthermore, the method for obtaining the fill light strategy includes:

[0050] The first step light requirement, the second fill light requirement and the area coordinates are combined into a fill light data set, and the fill light data set is divided into a training set and a test set;

[0051] Constructing a submerged plant intelligent lighting model, which includes an input layer, a hidden layer, a parameter selection layer and a data output layer;

[0052] The training set data is input into the input layer for preprocessing, the preprocessed data enters the hidden layer for feature extraction to obtain feature data, and the feature data is input into the parameter selection layer to select the fill light output fusion feature;

[0053] The parameter selection layer includes an attention layer and a feature fusion layer. The feature data enters the attention layer to weight the features, and the weighted features and the original features are input into the feature fusion layer for integration and output fusion features to the data output layer.

[0054] The data output layer consists of two fully connected layers. The fusion feature input is sent to the first fully connected layer and the output data is sent to the second fully connected layer. The second fully connected layer outputs the predicted value of the fill light parameter.

[0055] The root mean square error loss function is used to evaluate the difference between the predicted value and the true value of the lighting parameter, the Adam optimizer is used to optimize the model weight, and the test set is used to evaluate the model and output the intelligent lighting model for submerged plants.

[0056] The environmental data and environmental image of the area to be supplemented with light are input into the intelligent supplementary lighting model of submerged plants to obtain the supplementary lighting strategy and mobilize the supplementary lighting equipment for supplementary lighting.

[0057] The second aspect is the intelligent lighting system for submerged plants based on optical fiber, including:

[0058] Data acquisition module: used to obtain environmental images and environmental data of submerged plants, pre-process the environmental images and environmental data, and transmit them to the image module and the data module; the environmental images are obtained through a high-definition underwater camera, and the environmental data are obtained through a temperature sensor, a spectrum sensor, a dissolved oxygen meter, and a water quality carbon dioxide concentration meter;

[0059] Image module: used to segment the environmental image according to the growth area to obtain the regional environmental image, and perform image recognition on the regional environmental image to obtain the plant quantity, plant category, plant growth stage and water quality environment;

[0060] Data module: used to determine the regional lighting demand according to the lighting conditions, the number of plants and the lighting environment data, determine the first supplementary lighting demand according to the underwater temperature, the biological richness and the regional lighting demand, and determine the second supplementary lighting demand according to the oxygen content, the carbon dioxide content, the water quality environment and the regional lighting demand;

[0061] A fill light model module is used to construct a submerged plant intelligent fill light model according to the first fill light requirement, the second fill light requirement and the regional coordinates, and process the environmental data and environmental image of the area to be filled light and input them into the submerged plant intelligent fill light model to obtain fill light parameters; the fill light parameters include color temperature parameters, fill light intensity, fill light power and regional coordinates;

[0062] Intelligent supervision module: used for storing, viewing and managing the environmental image, the environmental data and the fill-light parameters, determining the fill-light strategy according to the fill-light parameters, and the controller sending adjustment signals of luminous intensity, color temperature and power according to the fill-light strategy; used for converting the submerged plant growth area into a cloud image through regional coordinates, viewing and managing the fill-light strategy of the submerged plant growth area with different regional coordinates;

[0063] Fill-light module: used to execute the adjustment signal of the controller to perform intelligent fill-light on the submerged plant growth area; comprises a light source group, an optical fiber transmission group, a fill-light group and a power supply group; the light source group consists of an optical fiber light source and RGB-LED lamp beads, and is used to change the luminous intensity, color temperature and power of the light source according to the adjustment signal; the optical fiber transmission group consists of an optical fiber bundle and an optical fiber connector, and is used to transmit the light source underwater; the fill-light group is used to evenly disperse the light source transmitted by the optical fiber to the submerged plant growth area for intelligent fill-light; the power supply group consists of a solar panel and a battery pack, and is used to provide power for the submerged plant intelligent fill-light system.

[0064] The beneficial effects of the present invention are:

[0065] The present invention is an intelligent light supplement system for submerged plants based on optical fiber and a control method thereof. Compared with the prior art, the present invention has the following technical effects:

[0066] The present invention can improve the accuracy of intelligent fill lighting through image segmentation, image recognition, determining regional lighting requirements, determining the first fill lighting requirement, determining the second fill lighting requirement and building a model, thereby improving the speed and precision of intelligent fill lighting for submerged plants, greatly saving resources, and improving fill lighting efficiency. It can realize the intelligent, efficient and remote controllable fill lighting for submerged plants, perform fill lighting control on submerged plants in real time, and provide guarantee for the timeliness and accuracy of intelligent fill lighting for submerged plants. It is of great significance to the intelligent fill lighting for submerged plants, can adapt to different intelligent fill lighting systems for submerged plants and the intelligent fill lighting requirements of submerged plants of different users, and has a certain universality. BRIEF DESCRIPTION OF THE DRAWINGS

[0067] Figure 1 The present invention is a flowchart of the steps of the optical fiber-based submerged plant intelligent lighting system and its control method. DETAILED DESCRIPTION

[0068] The present invention is further described below by means of specific embodiments. The illustrative embodiments and descriptions of the present invention are used to explain the present invention but are not intended to limit the present invention.

[0069] The optical fiber-based submerged plant intelligent lighting system and control method thereof of the present invention comprises the following steps:

[0070] like Figure 1 As shown, in this embodiment, the following steps are included:

[0071] Acquire environmental images and environmental data of submerged plants, and pre-process the environmental images and environmental data; the environmental data includes growth environment and lighting environment;

[0072] The environmental image is segmented according to the submerged plant growth area to obtain a regional environmental image, image recognition is performed on the regional environmental image to obtain an image recognition result, and the biological richness is calculated according to the image recognition result; the image recognition result includes the number of plants, plant types, plant growth stages and water quality environment;

[0073] Matching the plant category and the plant growth stage with a standard plant information database to obtain the lighting conditions of submerged plants, and determining the regional lighting requirements according to the lighting conditions, the number of plants and the lighting environment data;

[0074] Extracting features of the growth environment to obtain underwater temperature, oxygen content, and carbon dioxide content, determining a first supplemental lighting requirement according to the underwater temperature, the biological richness, and the regional lighting requirement, and determining a second supplemental lighting requirement according to the oxygen content, the carbon dioxide content, the water quality environment, and the regional lighting requirement;

[0075] A submerged plant intelligent fill-light model is constructed, the first fill-light requirement and the second fill-light requirement are input into the submerged plant intelligent fill-light model to obtain a fill-light strategy, and intelligent fill-light is performed on the submerged plants according to the fill-light strategy.

[0076] In this embodiment, the method for segmenting the environment image according to the submerged plant growth area to obtain the regional environment image includes:

[0077] The growth range of the submerged plants is projected onto a two-dimensional plane, the cell area is determined according to the fill light range of the optical fiber fill light device, and the growth area of ​​the submerged plants is divided according to the cell area;

[0078] The environmental image is segmented according to the submerged plant growth area to obtain a regional environmental image, including:

[0079] The Transformer model improved by the window multi-head attention mechanism is used for edge detection of the image. The ConvNext network is used to extract global and local features. The progressive upsampling decoder and the step-by-step aggregation module are used for feature fusion to complete image segmentation. The regional environment image is output and the regional coordinates are set for the regional environment image.

[0080] In the actual evaluation, taking the fiber optic intelligent lighting system for submerged plants in the XX Town Wetland Reserve as an example, the lighting range of a single fiber optic lighting device is 9㎡, and the corresponding regional environmental image shows the image content of the submerged plant growth range projected on a two-dimensional plane within 9㎡. The environmental image of the submerged plant growth area with the regional coordinates of (2, 3) is taken for image segmentation.

[0081] In this embodiment, the method for performing image recognition on the regional environment image includes:

[0082] (1) Construct a support vector machine image classification and counting model, obtain the color features, texture features and shape features in the regional environment image by color histogram, local binary pattern and contour analysis respectively, preprocess the extracted features, select radial basis function as kernel function to process data, use training set to train the model, use test set to evaluate the model performance, and output the support vector machine image classification and counting model;

[0083] Input the regional environmental image into the support vector machine image classification and counting model to output the classification and counting results of plants and animals in the regional environmental image, and record the categories and numbers of submerged plants;

[0084] In the actual evaluation, the regional environment image with the regional coordinates (2, 3) was input into the support vector machine image classification counting model for image recognition, and the output animal categories were 7 categories, the total number of animal individuals was 84 / species, the plant categories were 9 categories, and the total number of plant individuals was 251 / clumps. The submerged plants were: 100 Vallisneria, 30 clumps of Ceratophyllum, 20 Myriophyllum, 50 Hydrilla, and 40 Potamogeton.

[0085] (2) The phenological extraction method of time series NDVI is used to obtain the growth stage of submerged plants. The specific steps include:

[0086] Draw a region of interest enclosed by 3 or more points on the regional environment image, and calculate the average value of the three color bands in the region of interest:

[0087]

[0088] Where DN is the pixel brightness value of the red, blue and green bands, c is one of the red, blue and green bands, SUMD is the sum of all pixels of a certain band in the region of interest, and N is the number of pixels in the regional environment image;

[0089] Extract the vegetation index that characterizes the growth characteristics of submerged plants. The calculation formula of the vegetation index is:

[0090]

[0091] NDVI, GI, BI, RI, BRI are normalized vegetation index, relative greenness index, relative blueness index, relative redness index and relative blue-red brightness index, respectively. DN , G DN , R DN , B DN They are the near-infrared band pixel brightness value, green band pixel brightness value, red band pixel brightness value and blue band pixel brightness value respectively;

[0092] The inverse Fourier transform function ifft was used to reconstruct the time series NDVI, and the NDVI points within the interval were fitted to obtain the seasonal trajectory curve of submerged plant growth. The root mean square error was used to evaluate the fitting effect.

[0093] The growth season trajectory curve of submerged plants is expressed as:

[0094]

[0095] Where c1 is the mean value of light and vegetation fraction in winter, c2 is the difference between light and vegetation fraction in summer and winter, c3 is the trend of greenness in summer, t is time, x1 is the inflection point of greenness increase in spring, x3 is the inflection point of greenness decrease in autumn, x2 and x4 are parameters for adjusting the seasonal growth curves in spring and autumn, x2 and x4∈(0,3];

[0096] Extract phenological parameters according to the growth season trajectory curve of submerged plants to obtain plant growth stages and corresponding growth parameters; the plant growth stages include greening period, maturity period, leaf fall period and dormancy period;

[0097] In the actual evaluation, image recognition was performed on the regional environmental image with regional coordinates (2, 3) and the growth season trajectory curve of submerged plants was fitted, and the following were obtained: Vallisneria: 10 plants (greening period), 60 plants (mature period), 30 plants (leaf falling period); Ceratophyllum: 2 clumps (greening period), 15 clumps (mature period), 13 clumps (leaf falling period); Foxtail Algae: 4 plants (greening period), 12 plants (mature period), 4 plants (leaf falling period); Hydrilla: 15 plants (greening period), 30 plants (mature period), 5 plants (leaf falling period); Potamogeton: 7 plants (greening period), 24 plants (mature period), 9 plants (leaf falling period).

[0098] (3) Identify impurities in the water in the regional environment image to determine the water quality environment, identify impurities in the water in the regional environment image to determine the water quality environment, perform Gaussian filtering to denoise the regional environment image, and convert it into a grayscale image. Calculate the three Tamura texture features of the grayscale image, namely, medium roughness, directionality, and uniformity. Perform a comprehensive evaluation based on the set roughness threshold, directionality threshold, and uniformity threshold to obtain a water quality score. Combine the water quality score, pH value, and salt ion concentration to determine the water quality environment;

[0099] In the actual assessment, the water quality score of the regional environmental image with regional coordinates (2, 3) was 7 points after image recognition, the pH value was 7.5, the salt ion concentration was 110 mg / L, and the water quality environment was relatively good.

[0100] In this embodiment, the method for calculating biological richness according to the image recognition result includes:

[0101] Substitute the species category and species number obtained by image recognition into the biological richness function to obtain the biological richness. The expression of the biological richness function is:

[0102]

[0103] Where B is the biological richness, T is the number of all organisms in the area, C is the total number of all biological categories in the area, A is the area, and n is the total area. i 、n j 、n k are the number of aquatic plants of category i, aquatic animals of category j and microorganisms of category k, respectively; w1, w2, w3 are the category weights of aquatic plants, aquatic animals and microorganisms, respectively; l, m, n are the number of categories of aquatic plants, aquatic animals and microorganisms, respectively;

[0104] In the actual evaluation, the biological richness was calculated to be 7 based on the image recognition results of the regional environmental image with the regional coordinates (2, 3).

[0105] In this embodiment, the method for determining the regional lighting requirement includes:

[0106] According to the plant category and plant growth stage, the plant category and plant growth stage of the standard plant information database are matched, and the similarity between the submerged plants to be matched and the submerged plants in the standard plant information database is calculated one by one, and the light conditions of the submerged plants in the standard plant information database corresponding to the matching group with the highest similarity are extracted;

[0107] Lighting conditions include ideal spectral range and ideal light intensity. Feature extraction of lighting environment data is performed to obtain light intensity, spectral range, photon flux density, and radiation flux density. The ideal light intensity is adjusted according to the lighting environment data and the number of plants, and the regional light demand is calculated. The expression of regional light demand is:

[0108]

[0109] Among them I y ′ is the required light intensity in the area, α and β are the light parameter influencing factors, PPFD is the light quantum flux density, N is the number of all submerged plants in the area, l is the number of submerged plant categories, I yi is the ideal light intensity for the i-th type of submerged plants, PAR is the radiation flux density, w 1i is the weight coefficient of the i-th type of submerged plants, n i is the number of submerged plants of type i;

[0110] The spectral ranges of all submerged plants are sorted according to the value size, the largest starting wavelength is taken as the spectrum end judgment point, and the smallest ending wavelength is taken as the spectrum start judgment point. The regional required spectral range is determined according to the spectrum end judgment point and the spectrum start judgment point, and the regional required light intensity and the regional required spectral range are combined to determine the regional light demand;

[0111] In the actual evaluation, the illumination environment data of the area coordinates (2, 3) are: illumination intensity 5000 lux, spectral range 400-700 nm, photon flux density about 250 μmol / m 2 / s, combined with the ideal spectral range and ideal light intensity of submerged plants, when α is 0.8 and β is 1.1, the regional required light intensity is 7900 lux and the regional required spectral range is 350-890nm.

[0112] In this embodiment, the method for determining the first fill light requirement includes:

[0113] The environmental factors that affect the photosynthesis of submerged plants are determined based on the underwater temperature and biological richness. The underwater temperature, the ideal temperature of submerged plants and biological richness are substituted into the environmental function to obtain the environmental factors. The environmental function expression is:

[0114]

[0115] Where E is the environmental factor, x1 and x3 are the positive coefficients of the influence of the adjustment factor, T is the underwater temperature, and T y is the ideal temperature corresponding to the maximum efficiency of photosynthesis of submerged plants, x2 is the sensitivity of temperature change to environmental factors, and there is no linear relationship between environmental factors and the first term, B is biological richness, x4 is the sensitivity of biological richness to environmental factors, and the second term is the attenuation coefficient, and environmental factors are negatively correlated with it;

[0116] The first fill light requirement is determined according to environmental factors and regional lighting requirements. The corresponding expression of the lighting intensity of the first fill light requirement is:

[0117]

[0118] Where ΔI1 is the light intensity required for the first fill light, I y ′ is the required light intensity of the area, I t is the light intensity corresponding to the underwater lighting environment;

[0119] The spectral range required by the first fill light is consistent with the spectral range required by the area;

[0120] In the actual evaluation, T is taken as 20℃, T yTaking 22℃, x1 as 1.221, x2 as 0.1, x3 as 0.5, x4 as 0.1, combined with the biological richness B as 7, the environmental factor E is calculated to be 1.55;

[0121] Combining the regional required light intensity of 7900 lux and the light intensity of 5000 lux, it is calculated that the light intensity ΔI1 required for the first supplementary light is 4247 lux, and the spectral range of the first supplementary light is 350-890 nm.

[0122] In this embodiment, the method for determining the second fill light requirement includes:

[0123] Substituting the oxygen content, carbon dioxide content and water quality environment in the water into the second fill light demand function, the light intensity required for the second fill light is obtained. The expression of the second fill light demand function is:

[0124]

[0125] Among them, ΔI2 is the light intensity required for the second supplementary light, k1 and k2 are the influence coefficients of oxygen and carbon dioxide, d1 and d2 are the oxygen content and carbon dioxide content in water, r1 and r2 are the response coefficients of oxygen and carbon dioxide, p and q are the changes in oxygen content and carbon dioxide content, PH and PH y is the pH value in water and the ideal pH value, c, c y is the nutrient ion concentration and ideal nutrient ion concentration in water, o, o y to rate water quality and ideal water quality;

[0126] The spectral range required by the second fill light is consistent with the spectral range required by the area;

[0127] In the actual evaluation, the oxygen content and carbon dioxide content were 10 mg / L and 7 mg / L respectively, the corresponding influence coefficients k1 and k2 were both 1, the corresponding response coefficients r1 and r2 were 0.1 and 0.01 respectively, the pH value and the ideal pH value were both 7.5, and the nutrient ion concentration c in water and the ideal nutrient ion concentration c y is 110mg / L and 220mg / L, the water quality score and ideal water quality score are 7 and 7.5, and the value in the brackets is 0.87;

[0128] Combining the regional required light intensity of 7900 lux and the light intensity of 5000 lux, it is calculated that the light intensity ΔI1 required for the second fill light is 1873 lux, and the spectral range of the second fill light requirement is 350-890 nm.

[0129] In this embodiment, the method for obtaining the fill light strategy includes:

[0130] The first step light requirement, the second fill light requirement and the area coordinates are combined into a fill light data set, and the fill light data set is divided into a training set and a test set;

[0131] Constructing a submerged plant intelligent lighting model, which includes an input layer, a hidden layer, a parameter selection layer and a data output layer;

[0132] The training set data is input into the input layer for preprocessing, and the preprocessed data enters the hidden layer. The activation function uses ReLU to extract the data features to obtain feature data. The feature data is input into the parameter selection layer to select the fill light parameters; the feature data includes light intensity, maximum wavelength, minimum wavelength and regional coordinates;

[0133] The parameter selection layer includes the attention layer and the feature fusion layer. The feature data enters the attention layer, and the multi-layer perceptron MLP is used for parameter learning to calculate the attention weight. The activation function uses softmax, and the attention weight is used to weight the features. The weighted features and the original features are input into the feature fusion layer, and the features are integrated and output to the data output layer using ASFF.

[0134] The data output layer consists of two fully connected layers. The fused features are input into the first fully connected layer. The number of neurons in the first fully connected layer is set to 8, and the activation function adopts Leaky ReLU. The data is output to the second fully connected layer. The number of neurons in the second fully connected layer is set to 4. The tanh activation function is used to output continuous color temperature parameters, and the sigmoid activation function is used to output fill light intensity and fill light power. The data output layer outputs the predicted values ​​of fill light parameters including color temperature parameters, fill light intensity, fill light power and regional coordinates.

[0135] The root mean square error loss function was used to evaluate the difference between the predicted value and the true value of the fill-light parameter. The Adam optimizer was used to optimize the weights of the submerged plant intelligent fill-light model. The test set was used to evaluate the submerged plant intelligent fill-light model.

[0136] The environmental data and environmental image of the area to be supplemented with light are input into the intelligent supplementary lighting model of submerged plants to obtain supplementary lighting parameters, and the supplementary lighting strategy is determined according to the supplementary lighting parameters to mobilize the supplementary lighting equipment for supplementary lighting;

[0137] In the actual evaluation, the regional coordinates (2, 3), the first fill light requirement of 4247lux / 350-890nm, and the second fill light requirement of 1873lux / 350-890nm were input into the submerged plant intelligent fill light model to obtain specific fill light parameters: fill light intensity of 3297lux, fill light color temperature of 350-400nm and 700-890nm, and fill light power of 12w.

[0138] The second aspect is the intelligent lighting system for submerged plants based on optical fiber, including:

[0139] Data acquisition module: used to obtain environmental images and environmental data of submerged plants, pre-process the environmental images and environmental data, and transmit them to the image module and the data module; the environmental images are obtained through a high-definition underwater camera, and the environmental data are obtained through a temperature sensor, a spectrum sensor, a dissolved oxygen meter, and a water quality carbon dioxide concentration meter;

[0140] Image module: used to segment the environmental image according to the growth area to obtain the regional environmental image, and perform image recognition on the regional environmental image to obtain the plant quantity, plant category, plant growth stage and water quality environment;

[0141] Data module: used to determine the regional lighting demand according to the lighting conditions, the number of plants and the lighting environment data, determine the first supplementary lighting demand according to the underwater temperature, the biological richness and the regional lighting demand, and determine the second supplementary lighting demand according to the oxygen content, the carbon dioxide content, the water quality environment and the regional lighting demand;

[0142] A fill light model module is used to construct a submerged plant intelligent fill light model according to the first fill light requirement, the second fill light requirement and the regional coordinates, and process the environmental data and environmental image of the area to be filled light and input them into the submerged plant intelligent fill light model to obtain fill light parameters; the fill light parameters include color temperature parameters, fill light intensity, fill light power and regional coordinates;

[0143] Intelligent supervision module: used for storing, viewing and managing the environmental image, the environmental data and the fill-light parameters, determining the fill-light strategy according to the fill-light parameters, and the controller sending adjustment signals of luminous intensity, color temperature and power according to the fill-light strategy; used for converting the submerged plant growth area into a cloud image through regional coordinates, viewing and managing the fill-light strategy of the submerged plant growth area with different regional coordinates;

[0144] Fill-light module: used to execute the adjustment signal of the controller to perform intelligent fill-light on the submerged plant growth area; comprises a light source group, an optical fiber transmission group, a fill-light group and a power supply group; the light source group consists of an optical fiber light source and RGB-LED lamp beads, and is used to change the luminous intensity, color temperature and power of the light source according to the adjustment signal; the optical fiber transmission group consists of an optical fiber bundle and an optical fiber connector, and is used to transmit the light source underwater; the fill-light group is used to evenly disperse the light source transmitted by the optical fiber to the submerged plant growth area for intelligent fill-light; the power supply group consists of a solar panel and a battery pack, and is used to provide power for the submerged plant intelligent fill-light system.

[0145] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. An intelligent light supplement control method for submerged plants based on optical fiber, characterized in that: The following steps are involved: S1. Acquire environmental images and environmental data of submerged plants, and pre-process the environmental images and environmental data; the environmental data includes growth environment and lighting environment; S2. Segment the environmental image according to the submerged plant growth area to obtain a regional environmental image, perform image recognition on the regional environmental image to obtain an image recognition result, and calculate the biological richness according to the image recognition result; the image recognition result includes the number of plants, plant types, plant growth stages and water quality environment; S3, matching the plant category and the plant growth stage with a standard plant information database to obtain the lighting conditions of submerged plants, and determining the regional lighting requirements according to the lighting conditions, the number of plants and the lighting environment data; S4, extracting features of the growth environment to obtain underwater temperature, oxygen content and carbon dioxide content, determining a first supplementary lighting requirement according to the underwater temperature, the biological richness and the regional lighting requirement, and determining a second supplementary lighting requirement according to the oxygen content, the carbon dioxide content, the water quality environment and the regional lighting requirement; S5. Construct an intelligent supplementary lighting model for submerged plants, input the first supplementary lighting requirement and the second supplementary lighting requirement into the intelligent supplementary lighting model for submerged plants to obtain a supplementary lighting strategy, and perform intelligent supplementary lighting on the submerged plants according to the supplementary lighting strategy.

2. According to claim 1, the optical fiber-based intelligent light supplement control method for submerged plants is characterized in that: The method for obtaining a regional environmental image by segmenting the environmental image according to the submerged plant growth area comprises: The growth range of the submerged plants is projected onto a two-dimensional plane, the cell area is determined according to the fill light range of the optical fiber fill light device, and the growth area of ​​the submerged plants is divided according to the cell area; The environmental image is segmented according to the growth area of ​​the submerged plants to obtain a regional environmental image, and regional coordinates are set for the regional environmental image.

3. According to claim 1, the optical fiber-based intelligent light supplement control method for submerged plants is characterized in that: The method for performing image recognition on the regional environment image comprises: (1) Construct a support vector machine image classification and counting model, input the regional environmental image into the support vector machine image classification and counting model, output the classification and counting results of plants and animals in the regional environmental image, and record the types and numbers of submerged plants; (2) The phenological extraction method of time series NDVI is used to obtain the growth stages of submerged plants, including: Draw a region of interest enclosed by 3 or more points on the regional environment image, and calculate the average values ​​of the three color bands in the region of interest; Extract vegetation indices that characterize the growth characteristics of submerged plants, including the normalized difference vegetation index NDVI, relative greenness index GI, relative blueness index BI, relative redness index RI, and relative blue-red brightness BRI index; The inverse Fourier transform function ifft was used to reconstruct the time series NDVI, and the NDVI points within the interval were fitted to obtain the seasonal trajectory curve of submerged plant growth. The root mean square error was used to evaluate the fitting effect. The growth season trajectory curve of submerged plants is expressed as: Where c1 is the mean value of light and vegetation fraction in winter, c2 is the difference between light and vegetation fraction in summer and winter, c3 is the trend of greenness in summer, t is time, x1 is the inflection point of greenness increase in spring, x3 is the inflection point of greenness decrease in autumn, x2 and x4 are parameters for adjusting the seasonal growth curves in spring and autumn, x2 and x4∈(0,3]; According to the seasonal trajectory curve of submerged plant growth, phenological parameters are extracted to obtain plant growth stages and corresponding growth parameters; (3) Identify impurities in the water in the regional environmental image to determine the water quality environment, perform Gaussian filtering to denoise the regional environmental image, and convert it into a grayscale image. Calculate the three Tamura texture features of the grayscale image, namely, medium roughness, directionality, and uniformity. Perform a comprehensive evaluation based on the set roughness threshold, directionality threshold, and uniformity threshold to obtain a water quality score. Combine the water quality score, pH value, and salt ion concentration to determine the water quality environment.

4. According to the optical fiber-based intelligent supplementary lighting control method for submerged plants in claim 1, it is characterized in that: The method for calculating biological richness according to the image recognition result comprises: Substitute the species category and species number obtained by image recognition into the biological richness function to obtain the biological richness. The expression of the biological richness function is: Where B is the biological richness, T is the number of all organisms in the area, C is the total number of all biological categories in the area, A is the area, and n is the total area. i 、n j 、n k are the number of aquatic plants of the ith category, aquatic animals of the jth category and microorganisms of the kth category, w1, w2, w3 are the category weights of aquatic plants, aquatic animals and microorganisms, l, m, n are the number of categories of aquatic plants, aquatic animals and microorganisms, respectively.

5. According to claim 1, the optical fiber-based intelligent supplementary lighting control method for submerged plants is characterized in that: The method for determining the lighting requirement of the area comprises: According to the plant category and plant growth stage, the plant category and plant growth stage of the standard plant information database are matched, and the similarity between the submerged plants to be matched and the submerged plants in the standard plant information database is calculated one by one, and the light conditions of the submerged plants in the standard plant information database corresponding to the matching group with the highest similarity are extracted; Lighting conditions include ideal spectral range and ideal light intensity. Feature extraction of lighting environment data is performed to obtain light intensity, spectral range, photon flux density, and radiation flux density. The ideal light intensity is adjusted according to the lighting environment data and the number of plants, and the regional light demand is calculated. The expression of regional light demand is: Among them I y ′ is the required light intensity in the area, α and β are the light parameter influencing factors, PPFD is the light quantum flux density, N is the number of all submerged plants in the area, l is the number of submerged plant categories, I yi is the ideal light intensity for the i-th type of submerged plants, PAR is the radiation flux density, w 1i is the weight coefficient of the i-th type of submerged plants, n i is the number of submerged plants of type i; The spectral ranges of all submerged plants are sorted according to the size of the values, the largest starting wavelength is taken as the spectrum end judgment point, and the smallest ending wavelength is taken as the spectrum starting judgment point. The regional required spectral range is determined based on the spectrum end judgment point and the spectrum starting judgment point, and the regional required light intensity and the regional required spectral range are combined to determine the regional light demand.

6. According to the optical fiber-based intelligent light supplement control method for submerged plants in claim 1, it is characterized in that: The method for determining the first fill light requirement includes: The environmental factors that affect the photosynthesis of submerged plants are determined based on the underwater temperature and biological richness. The underwater temperature, the ideal temperature of submerged plants and biological richness are substituted into the environmental function to obtain the environmental factors. The environmental function expression is: Where E is the environmental factor, x1 and x3 are the positive coefficients of the influence of the adjustment factor, T is the underwater temperature, and T y is the ideal temperature corresponding to the maximum efficiency of photosynthesis of submerged plants, x2 is the sensitivity of temperature change to environmental factors, and there is no linear relationship between environmental factors and the first term, B is biological richness, x4 is the sensitivity of biological richness to environmental factors, and the second term is the attenuation coefficient, and environmental factors are negatively correlated with it; The first fill light requirement is determined according to environmental factors and regional lighting requirements. The corresponding expression of the lighting intensity of the first fill light requirement is: Where ΔI1 is the light intensity required for the first fill light, I′ y is the required light intensity of the area, I t is the light intensity corresponding to the underwater lighting environment; The spectral range required by the first fill light is consistent with the spectral range required by the area.

7. The optical fiber-based intelligent light supplement control method for submerged plants according to claim 1, characterized in that: The method for determining the second fill light requirement includes: Substituting the oxygen content, carbon dioxide content and water quality environment in the water into the second fill light demand function, the light intensity required for the second fill light is obtained. The expression of the second fill light demand function is: Among them, ΔI2 is the light intensity required for the second supplementary light, k1 and k2 are the influence coefficients of oxygen and carbon dioxide, d1 and d2 are the oxygen content and carbon dioxide content in water, r1 and r2 are the response coefficients of oxygen and carbon dioxide, p and q are the changes in oxygen content and carbon dioxide content, PH and PH y is the pH value in water and the ideal pH value, c, c y is the nutrient ion concentration and ideal nutrient ion concentration in water, o, o y to rate water quality and ideal water quality; The spectral range required by the second fill light is consistent with the spectral range required by the area.

8. The optical fiber-based intelligent supplementary lighting control method for submerged plants according to claim 1, characterized in that: The method for obtaining the fill light strategy includes: The first step light requirement, the second fill light requirement and the area coordinates are combined into a fill light data set, and the fill light data set is divided into a training set and a test set; Constructing a submerged plant intelligent lighting model, which includes an input layer, a hidden layer, a parameter selection layer and a data output layer; The training set data is input into the input layer for preprocessing, the preprocessed data enters the hidden layer for feature extraction to obtain feature data, and the feature data is input into the parameter selection layer to select the fill light output fusion feature; The parameter selection layer includes an attention layer and a feature fusion layer. The feature data enters the attention layer to weight the features, and the weighted features and the original features are input into the feature fusion layer for integration and output fusion features to the data output layer. The data output layer consists of two fully connected layers. The fusion feature input is sent to the first fully connected layer and the output data is sent to the second fully connected layer. The second fully connected layer outputs the predicted value of the fill light parameter. The root mean square error loss function was used to evaluate the difference between the predicted value and the true value of the fill-light parameter. The Adam optimizer was used to optimize the weights of the submerged plant intelligent fill-light model. The test set was used to evaluate the submerged plant intelligent fill-light model. The environmental data and environmental image of the area to be supplemented with light are input into the intelligent supplementary lighting model of submerged plants to obtain the supplementary lighting strategy and mobilize the supplementary lighting equipment for supplementary lighting.

9. An optical fiber-based intelligent lighting system for submerged plants, used to implement the method of claims 1-8, characterized in that: include: Data acquisition module: used to obtain environmental images and environmental data of submerged plants, pre-process the environmental images and environmental data, and transmit them to the image module and the data module; the environmental images are obtained through a high-definition underwater camera, and the environmental data are obtained through a temperature sensor, a spectrum sensor, a dissolved oxygen meter, and a water quality carbon dioxide concentration meter; Image module: used to segment the environmental image according to the growth area to obtain the regional environmental image, and perform image recognition on the regional environmental image to obtain the plant quantity, plant category, plant growth stage and water quality environment; Data module: used to determine the regional lighting demand according to the lighting conditions, the number of plants and the lighting environment data, determine the first supplementary lighting demand according to the underwater temperature, the biological richness and the regional lighting demand, and determine the second supplementary lighting demand according to the oxygen content, the carbon dioxide content, the water quality environment and the regional lighting demand; A fill light model module is used to construct a submerged plant intelligent fill light model according to the first fill light requirement, the second fill light requirement and the regional coordinates, and process the environmental data and environmental image of the area to be filled light and input them into the submerged plant intelligent fill light model to obtain fill light parameters; the fill light parameters include color temperature parameters, fill light intensity, fill light power and regional coordinates; Intelligent supervision module: used for storing, viewing and managing the environmental image, the environmental data and the fill-light parameters, determining the fill-light strategy according to the fill-light parameters, and the controller sending adjustment signals of luminous intensity, color temperature and power according to the fill-light strategy; used for converting the submerged plant growth area into a cloud image through regional coordinates, viewing and managing the fill-light strategy of the submerged plant growth area with different regional coordinates; Fill-light module: used to execute the adjustment signal of the controller to perform intelligent fill-light on the submerged plant growth area; comprises a light source group, an optical fiber transmission group, a fill-light group and a power supply group; the light source group consists of an optical fiber light source and RGB-LED lamp beads, and is used to change the luminous intensity, color temperature and power of the light source according to the adjustment signal; the optical fiber transmission group consists of an optical fiber bundle and an optical fiber connector, and is used to transmit the light source underwater; the fill-light group is used to evenly disperse the light source transmitted by the optical fiber to the submerged plant growth area for intelligent fill-light; the power supply group consists of a solar panel and a battery pack, and is used to provide power for the submerged plant intelligent fill-light system.

Images