Soil environment quality adjusting system and method for flower planting

Abstract

The invention discloses a soil environment quality adjusting system and method for flower planting, and relates to the field of plant irrigation, and the method comprises the steps: dividing a collection region, obtaining basic environment characteristics, layering soil according to the vertical depth, arranging environment sensors, and constructing a reference correlation model of each layer based on environment parameters when no flower exists; calling the model in a planting area, constructing a flower influence model by combining environmental parameters with flowers and layered soil water content, and calculating actual water consumption of each layer by comparing the two models and combining soil volume and volume weight; then analyzing a water consumption time sequence, dividing a stable water consumption section and a transition section in combination with a flower image shot by a camera, and calculating a water consumption coefficient to construct a layered prediction water consumption model; and finally, calling the model according to the to-be-predicted region identifier and the flower variety, calculating the predicted water consumption and alternately irrigating in a layered manner. The water consumption demand can be accurately matched, water resource waste and growth water shortage are avoided, and the growth stability, quality and survival rate of flowers are improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of plant irrigation, in particular to a soil environment quality regulation system and method for flower planting. BACKGROUND

[0002] Surface soil moisture is the basic guarantee for the growth and development of flowers, and has a significant regulatory effect on the formation, transformation and consumption process of land surface water resources. If the soil moisture is continuously reduced for a period of time, it will lead to flower growth obstruction, flower quality decline, and even leaf yellowing, flower bud shedding and other problems. Therefore, soil moisture is a key indicator for measuring the degree of drought in flower planting. Therefore, how to improve the monitoring accuracy of soil moisture content is crucial for accurately judging the degree of drought in flower planting and ensuring normal growth of flowers.

[0003] The current flower irrigation scheme has obvious limitations. The irrigation mode completely depends on manual operation, and the time and water volume of irrigation are adjusted by subjective judgment. Or it can only run according to the pre-set fixed rules, such as fixed time interval and unified water volume standard to execute irrigation action. No matter which way, it cannot realize flexible adjustment according to the different water demand characteristics of different growth stages of flowers, and cannot respond to real-time water demand changes caused by environmental changes, and it is always difficult to achieve the core goal of demand irrigation. SUMMARY

[0004] The purpose of the present application is to provide a soil environment quality regulation system and method for flower planting to solve the problems in the prior art.

[0005] To achieve the above purpose, the present application provides the following technical scheme: a soil environment quality regulation method for flower planting, the method comprising the following steps: Step S1, divide the collection area to obtain the basic environment characteristics in the collection area; divide the soil in the collection area into several layers, arrange the environment sensor, obtain the environment parameters in the collection area and obtain the soil volume water content without flowers, and construct the baseline correlation model of each layer; Further, step S1 comprises: Step S1-1, divide the areas with the same climate and terrain into the same collection area, and assign a unique area identifier to each collection area; the basic environment characteristics in the collection area include climate characteristics, terrain characteristics and environment characteristics; wherein the climate includes tropical rainforest climate, tropical grassland climate, tropical desert climate, tropical monsoon climate, subtropical monsoon climate and monsoon humid climate, Mediterranean climate, temperate marine climate, temperate continental climate, temperate monsoon climate, mountain climate, polar tundra climate and polar ice climate; the terrain includes plateau, basin, plain, hilly and mountainous area; The climate features include the annual average atmospheric humidity and the annual average atmospheric temperature range; the topographic features include altitude, slope and soil type; the environmental features include the annual average sunshine duration and wind speed. Step S1-2: Divide the microclimate zone within the collection area according to the basic environmental characteristics. Within each microclimate zone of the collection area, divide the detection area according to soil type and assign a sub-identifier to each detection area. The soil within the collection area is divided into several layers according to different vertical depths. Environmental sensors are deployed within the detection area, including humidity sensors, temperature sensors, light sensors, cup anemometers, and soil moisture meters. The environmental parameters include atmospheric humidity, atmospheric temperature, light intensity, and wind speed. The soil moisture meters are buried in layers. Steps S1-3: Construct a baseline correlation model for each layer of the detection area based on environmental parameters: S 0，j =aH+bT+cL+dV, where S represents the soil volumetric water content, S 0，j The soil volumetric water content of layer j without flowers is represented by H, atmospheric humidity by T, atmospheric temperature by L, light intensity by V, and wind speed by V. a, b, c, and d are weighting coefficients. The region identifier, sub-identifier, and environmental parameters are associated with the baseline association model of each layer and uploaded to the cloud.

[0006] Step S2: Obtain the planting area of ​​the flowers and call the baseline association model of each layer; obtain the environmental parameters of the planting area through environmental sensors and obtain the soil volume water content when there are flowers in each layer; obtain the actual water consumption of the flowers in each layer according to the baseline association model. Furthermore, step S2 includes: The planting area is a soil environment with physical boundaries and a constant volume within the boundaries; different flowers in the planting area are of the same variety; the area identifier and sub-identifier of the planting area are obtained, and the baseline association model associated with the cloud is called. Environmental sensors were deployed within the planting area to acquire environmental parameters, and the soil volumetric water content at each layer when flowers were present was obtained. A flower influence model was constructed for each layer according to the flower variety: S 1，j =eH+fT+gL+hV, where S represents the soil volumetric water content when flowers are present. 1，j denoted by , where represents the volumetric water content of the soil in the j-th layer when there are flowers, H represents atmospheric humidity, T represents atmospheric temperature, L represents light intensity, V represents wind speed, and e, f, g, and h are weighting coefficients, respectively. Based on the baseline correlation model and the flower impact model, the unit water consumption of flowers at each layer under the same environmental parameters is calculated as: ΔS j =S 0，j -S 1，j=(ae)H+(bf)T+(cg)L+(dh)V, ΔS j The water consumption per unit of flowers in the j-th layer is expressed as a percentage; the soil volume of the planting area is obtained, and the actual water consumption of flowers in different layers is calculated based on the soil volume: M j =ΔS j ×V s ×ρ, where M represents the actual water consumption of the flowers, M j ΔS represents the actual water consumption of the flowers in the j-th layer. j V represents the water consumption per unit of flowers in the j-th layer. s ρ represents the soil volume evenly distributed according to the number of layers within the planting area, and ρ represents the soil bulk density of the planting area; and records the timestamp of water consumption corresponding to the actual water consumption of flowers in each layer.

[0007] Step S3: Divide the actual water consumption of flowers in each layer into stable water consumption section and transition section, calculate the stable water consumption coefficient of the stable water consumption section and the dynamic water consumption coefficient of the transition section, and construct the predicted water consumption model for each layer. Furthermore, step S3 includes: Step S3-1: Set up a camera in the planting area, capture images of the flowers at a preset acquisition frequency, and record the image timestamps; Step S3-2: Extract flower varieties and flower features from the image through the convolutional layer of the convolutional neural network. The flower features include leaf texture features, the pixel ratio of flowers in the flower image, and flower coverage. The leaf texture features are the quantified values ​​of leaf texture complexity. The pixel ratio of flowers in the flower image is the ratio of the number of flower pixels to the total number of flower pixels in the image. The flower coverage is the percentage of the vertical projection area of ​​the flowers on the ground to the surface area of ​​the planting area. Step S3-3: Form a water consumption time series of the actual water consumption of flowers in each layer according to the order of water consumption timestamps; set a sliding window with a step size of n days, where n is an odd number; calculate the daily average change rate of the actual water consumption of flowers within the sliding window; when the daily average change rate of water consumption exceeds a preset threshold, mark the date of the (n-1) / 2th day from the start date of the sliding window as the change point; The layer closest to the soil surface is designated as the first layer. The water consumption time series of the first layer is divided into m segments according to the order of water consumption timestamps. The captured images are clustered using K-means clustering, and the images are clustered into m clusters according to the order of image timestamps. If the timestamps of the m segments do not match the timestamps of the corresponding m clusters within a preset timestamp range, an early warning is issued. If they match, the water consumption time series of each layer is divided into several water consumption segments based on the change points. Step S3-4: Divide the T days before and after the change point into transition segments. The T days before the transition segment belong to the stable water consumption segment A, and the T days after the transition segment belong to the stable water consumption segment B. And so on, divide the water consumption segments according to the stable water consumption segment and the transition segment. Calculate the average daily water consumption within each stable water consumption segment, and take the stable water consumption segment with the lowest average daily water consumption as the baseline segment. Define the stable water consumption coefficient K0 = 1.0 for the baseline segment; calculate the stable water consumption coefficient for each stable water consumption segment: ; Where K i M represents the steady water consumption coefficient of the i-th steady water consumption segment. i M0 represents the average daily water consumption of the i-th stable water consumption segment, and `M0 represents the average daily water consumption of the baseline segment. The dynamic water consumption coefficient of the transition section per day was calculated using linear interpolation. ; Where K(t) represents the dynamic water consumption coefficient on day t of the transition period, K A K represents the steady water consumption coefficient of steady water consumption section A. B The stable water consumption coefficient represents the stable water consumption section B; Step S3-5: Based on the stable water consumption coefficient and the dynamic water consumption coefficient, construct predictive water consumption models for different layers using flower varieties as a benchmark: M y，j =ΔS j ×V s ×ρ×K, where M y，j ΔS represents the predicted water consumption of flowers in the j-th layer; j V represents the water consumption per unit of flowers in the j-th layer; s ρ represents the soil volume evenly distributed according to the number of layers within the planting area; K represents the soil bulk density of the planting area; K represents the stable water consumption coefficient or dynamic water consumption coefficient of the j-th layer corresponding to the prediction date; the prediction water consumption models of different layers are stored in the cloud according to the region identifier, sub-identifier, cluster and flower variety.

[0008] Step S4: Obtain the varieties of flowers in the planting area to be predicted, output the predicted water consumption of each layer of flowers according to the predicted water consumption model of each layer, and irrigate the flowers in layers alternately.

[0009] Furthermore, step S4 includes: Step S4-1: Obtain the region identifier and sub-identifier of the planting area to be predicted, and retrieve the corresponding prediction water consumption model of different layers according to the flower variety; take actual images of flowers in the planting area to be predicted by camera, compare the similarity of the actual images with the images in the corresponding clusters, and take the image in the cluster with the highest similarity as the matching image. Step S4-2: Obtain the water consumption segment of each layer matched by the matching image, and obtain the stable water consumption segment or transition segment to which the matching image belongs based on the position of the water consumption segment in the water consumption time series of the corresponding layer; obtain the environmental parameters of the planting area to be predicted, obtain the unit water consumption of flowers in different layers based on the environmental parameters, calculate the predicted water consumption of the flower variety based on the predicted water consumption model, and perform layered alternating irrigation for layers where the predicted water consumption of flowers is not zero according to the preset time.

[0010] A soil environment quality regulation system for flower cultivation includes an environmental modeling module, a water consumption analysis module, a model training module, and an irrigation execution module. The environmental modeling module is used to construct a baseline correlation model for each layer. The water consumption analysis module is used to calculate the actual water consumption of flowers in each layer within the planting area. The model training module is used to construct a predicted water consumption model for each layer. The irrigation execution module is used to obtain the predicted water consumption of flowers in each layer and perform stratified alternating irrigation. The output of the environmental modeling module is connected to the input of the water consumption analysis module; the output of the water consumption analysis module is connected to the input of the model training module; and the output of the model training module is connected to the input of the irrigation execution module.

[0011] The environmental modeling module also includes a partitioning unit and a construction unit; the partitioning unit is used to divide the collection area and the detection area according to the climate characteristics and terrain characteristics, and to obtain basic environmental characteristics; the construction unit is used to deploy environmental sensors and to construct a baseline correlation model for each layer based on environmental parameters.

[0012] The water consumption analysis module also includes a calling unit and a calculation unit; the calling unit is used to call the baseline correlation model of each layer and obtain the environmental parameters within the planting area; the calculation unit is used to calculate the unit water consumption of flowers in each layer and the actual water consumption of flowers in each layer.

[0013] The model training module also includes a segmentation unit and a coefficient unit; the segmentation unit is used to capture images of flowers through a camera, extract flower features, and divide the water consumption time series of each layer into a stable water consumption segment and a transition segment; the coefficient unit is used to calculate the stable water consumption coefficient and the dynamic water consumption coefficient, and construct the predicted water consumption model for each layer.

[0014] The irrigation execution module further includes a matching unit and an execution unit; the matching unit is used to compare the similarity between the actual image captured by the camera and the image in the cluster and obtain the matching image; the execution unit is used to calculate the predicted water consumption of each layer according to environmental parameters and perform layered alternating irrigation.

[0015] Compared with the prior art, the beneficial effects of the present invention are: 1. This invention divides the soil into vertical layers within the collection area and deploys environmental sensors in each layer. First, a baseline correlation model for each layer is constructed based on environmental parameters when there are no flowers. Then, a flower impact model is constructed within the planting area, combining environmental parameters when flowers are present with soil volumetric water content. By comparing the two models, the unit water consumption of flowers in each layer is calculated, and the actual water consumption is obtained by combining the soil volume and bulk density of the planting area. This layered, precise calculation method completely solves the problem of water consumption estimation errors caused by neglecting soil moisture differences in traditional irrigation. It avoids water waste due to over-irrigation and prevents insufficient water supply from affecting flower growth, achieving precise matching of flower water consumption needs. This provides scientific data support for soil environmental water regulation, ensuring that flowers are always in a suitable water environment and effectively improving flower growth stability.

[0016] 2. This invention analyzes the time series of actual water consumption of each layer of flowers by setting a sliding window with odd-numbered days as the step size. Change points are marked when the daily average water consumption change rate exceeds a preset threshold. Simultaneously, images of the flowers are captured by a camera at a preset frequency. Key features such as leaf texture, flower pixel ratio, and flower coverage are extracted using a convolutional neural network. K-means clustering is then used to match the images with the timestamps of the water consumption segments, thereby dividing the water consumption segment into stable and transitional segments. Stable and dynamic water consumption coefficients are calculated separately, ultimately constructing a hierarchical predictive water consumption model. This dynamic adaptation mechanism can respond in real time to the water consumption changes of flowers at different growth stages, avoiding the shortcomings of traditional fixed irrigation methods that cannot adapt to the differences in water consumption throughout the growth cycle. It consistently maintains soil moisture within a range that matches the current growth needs of the flowers, effectively stabilizing environmental indicators such as soil moisture and nutrients, and significantly improving the growth quality and survival rate of the flowers.

[0017] 3. This invention obtains the regional identifier, sub-identifier, and flower variety of the planting area to be predicted, retrieves the corresponding stratified water consumption prediction model stored in the cloud, and compares the similarity between actual flower images captured by cameras and cluster images to determine the current stable water consumption segment or transition segment. Then, based on real-time environmental parameters, it calculates the predicted water consumption for each layer and implements stratified alternating irrigation for soil layers with non-zero predicted water consumption. Compared to traditional whole-soil irrigation methods, this strategy can accurately match the different water requirements of different soil layers—for example, shallow root concentration areas of flowers require less water per irrigation, while deep root areas require more. It avoids the problems of surface soil compaction due to over-irrigation and deep soil drought due to insufficient water replenishment, optimizing the uniformity of vertical soil water distribution. Simultaneously, it reduces soil nutrient loss with surface water accumulation, improves soil water and fertilizer retention capacity, maintains good soil physical structure and fertility levels in the long term, provides a suitable soil environment for flower root development, further promotes deep root growth and nutrient absorption, and achieves a synergistic improvement in soil environmental quality and flower growth status. Attached Figure Description

[0018] Figure 1 This is a schematic flowchart of a method for regulating soil environmental quality for flower cultivation according to the present invention. Figure 2 This is a schematic diagram of the structure of a soil environmental quality regulation system for flower planting according to the present invention. Detailed Implementation

[0019] Based on the embodiments of this invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this invention.

[0020] Example 1: As Figure 1 As shown, the present invention provides a technical solution: a method for regulating soil environmental quality for flower cultivation, the method comprising the following steps: Step S1: Divide the collection area and obtain the basic environmental characteristics within the collection area; divide the soil in the collection area into several layers, deploy environmental sensors, obtain environmental parameters in the collection area, and obtain the soil volumetric water content without flowers in each layer, and construct a baseline correlation model for each layer. Furthermore, step S1 includes: Step S1-1: Divide areas with similar climate and topography into the same data collection area, and assign a unique area identifier to each data collection area; the basic environmental characteristics within the data collection area include climate characteristics, topographic characteristics, and environmental characteristics; The climate features include the annual average atmospheric humidity and the annual average atmospheric temperature range; the topographic features include altitude, slope and soil type; the environmental features include the annual average sunshine duration and wind speed. Step S1-2: Divide the microclimate zone within the collection area according to the basic environmental characteristics. Within each microclimate zone of the collection area, divide the detection area according to soil type and assign a sub-identifier to each detection area. The soil within the collection area is divided into several layers according to different vertical depths. Environmental sensors are deployed within the detection area, including humidity sensors, temperature sensors, light sensors, cup anemometers, and soil moisture meters. The environmental parameters include atmospheric humidity, atmospheric temperature, light intensity, and wind speed. The soil moisture meters are buried in layers. Steps S1-3: Construct a baseline correlation model for each layer of the detection area based on environmental parameters: S 0，j =aH+bT+cL+dV, where S represents the soil volumetric water content, S 0，j The soil volumetric water content of layer j without flowers is represented by H, atmospheric humidity by T, atmospheric temperature by L, light intensity by V, and wind speed by V. a, b, c, and d are weighting coefficients. The region identifier, sub-identifier, and environmental parameters are associated with the baseline association model of each layer and uploaded to the cloud.

[0021] Step S2: Obtain the planting area of ​​the flowers and call the baseline association model of each layer; obtain the environmental parameters of the planting area through environmental sensors and obtain the soil volume water content when there are flowers in each layer; obtain the actual water consumption of the flowers in each layer according to the baseline association model. Furthermore, step S2 includes: The planting area is a soil environment with physical boundaries and a constant volume within the boundaries; different flowers in the planting area are of the same variety; the area identifier and sub-identifier of the planting area are obtained, and the baseline association model associated with the cloud is called. Environmental sensors were deployed within the planting area to acquire environmental parameters, and the soil volumetric water content at each layer when flowers were present was obtained. A flower influence model was constructed for each layer according to the flower variety: S 1，j =eH+fT+gL+hV, where S represents the soil volumetric water content when flowers are present. 1，j denoted by , where represents the volumetric water content of the soil in the j-th layer when there are flowers, H represents atmospheric humidity, T represents atmospheric temperature, L represents light intensity, V represents wind speed, and e, f, g, and h are weighting coefficients, respectively. Based on the baseline correlation model and the flower impact model, the unit water consumption of flowers at each layer under the same environmental parameters is calculated as: ΔS j =S 0，j -S 1，j =(ae)H+(bf)T+(cg)L+(dh)V, ΔS j The water consumption per unit of flowers in the j-th layer is expressed as a percentage; the soil volume of the planting area is obtained, and the actual water consumption of flowers in different layers is calculated based on the soil volume: M j =ΔS j ×V s ×ρ, where M represents the actual water consumption of the flowers, M j ΔS represents the actual water consumption of the flowers in the j-th layer. j V represents the water consumption per unit of flowers in the j-th layer. s ρ represents the soil volume evenly distributed according to the number of layers within the planting area, and ρ represents the soil bulk density of the planting area; and records the timestamp of water consumption corresponding to the actual water consumption of flowers in each layer.

[0022] Step S3: Divide the actual water consumption of flowers in each layer into stable water consumption section and transition section, calculate the stable water consumption coefficient of the stable water consumption section and the dynamic water consumption coefficient of the transition section, and construct the predicted water consumption model for each layer. Furthermore, step S3 includes: Step S3-1: Set up a camera in the planting area, capture images of the flowers at a preset acquisition frequency, and record the image timestamps; Step S3-2: Extract flower varieties and flower features from the image through the convolutional layer of the convolutional neural network. The flower features include leaf texture features, the pixel ratio of flowers in the flower image, and flower coverage. The leaf texture features are the quantified values ​​of leaf texture complexity. The pixel ratio of flowers in the flower image is the ratio of the number of flower pixels to the total number of flower pixels in the image. The flower coverage is the percentage of the vertical projection area of ​​the flowers on the ground to the surface area of ​​the planting area. Step S3-3: Form a water consumption time series of the actual water consumption of flowers in each layer according to the order of water consumption timestamps; set a sliding window with a step size of n days, where n is an odd number; calculate the daily average change rate of the actual water consumption of flowers within the sliding window; when the daily average change rate of water consumption exceeds a preset threshold, mark the date of the (n-1) / 2th day from the start date of the sliding window as the change point; The layer closest to the soil surface is designated as the first layer. The water consumption time series of the first layer is divided into m segments according to the order of water consumption timestamps. The captured images are clustered using K-means clustering, and the images are clustered into m clusters according to the order of image timestamps. If the timestamps of the m segments do not match the timestamps of the corresponding m clusters within a preset timestamp range, an early warning is issued. If they match, the water consumption time series of each layer is divided into several water consumption segments based on the change points. Step S3-4: Divide the T days before and after the change point into transition segments. The T days before the transition segment belong to the stable water consumption segment A, and the T days after the transition segment belong to the stable water consumption segment B. And so on, divide the water consumption segments according to the stable water consumption segment and the transition segment. Calculate the average daily water consumption within each stable water consumption segment, and take the stable water consumption segment with the lowest average daily water consumption as the baseline segment. Define the stable water consumption coefficient K0 = 1.0 for the baseline segment; calculate the stable water consumption coefficient for each stable water consumption segment: ; Where K i M represents the steady water consumption coefficient of the i-th steady water consumption segment. i M0 represents the average daily water consumption of the i-th stable water consumption segment, and `M0 represents the average daily water consumption of the baseline segment. The dynamic water consumption coefficient of the transition section per day was calculated using linear interpolation. ; Where K(t) represents the dynamic water consumption coefficient on day t of the transition period, K A K represents the steady water consumption coefficient of steady water consumption section A. B The stable water consumption coefficient represents the stable water consumption section B; Step S3-5: Based on the stable water consumption coefficient and the dynamic water consumption coefficient, construct predictive water consumption models for different layers using flower varieties as a benchmark: M y，j =ΔSj ×V s ×ρ×K, where M y，j ΔS represents the predicted water consumption of flowers in the j-th layer; j V represents the water consumption per unit of flowers in the j-th layer; s ρ represents the soil volume evenly distributed according to the number of layers within the planting area; K represents the soil bulk density of the planting area; K represents the stable water consumption coefficient or dynamic water consumption coefficient of the j-th layer corresponding to the prediction date; the prediction water consumption models of different layers are stored in the cloud according to the region identifier, sub-identifier, cluster and flower variety.

[0023] Step S4: Obtain the varieties of flowers in the planting area to be predicted, output the predicted water consumption of each layer of flowers according to the predicted water consumption model of each layer, and irrigate the flowers in layers alternately.

[0024] Furthermore, step S4 includes: Step S4-1: Obtain the region identifier and sub-identifier of the planting area to be predicted, and retrieve the corresponding prediction water consumption model of different layers according to the flower variety; take actual images of flowers in the planting area to be predicted by camera, compare the similarity of the actual images with the images in the corresponding clusters, and take the image in the cluster with the highest similarity as the matching image. Step S4-2: Obtain the water consumption segment of each layer matched by the matching image, and obtain the stable water consumption segment or transition segment to which the matching image belongs based on the position of the water consumption segment in the water consumption time series of the corresponding layer; obtain the environmental parameters of the planting area to be predicted, obtain the unit water consumption of flowers in different layers based on the environmental parameters, calculate the predicted water consumption of the flower variety based on the predicted water consumption model, and perform layered alternating irrigation for layers where the predicted water consumption of flowers is not zero according to the preset time.

[0025] For example: The area with a subtropical monsoon climate and plain terrain was designated as a single collection area, and the area was assigned the identifier HZ-001.

[0026] The basic environmental characteristics of the data collection area are as follows: Climate characteristics: Annual average atmospheric humidity 70%; The average annual atmospheric temperature ranges from 15℃ to 28℃. Topographic features: elevation 50 meters, slope 2°, soil type: loam; The environmental characteristics include an average annual sunshine duration of 1800 hours and an average annual wind speed of 2.5 meters per second.

[0027] Based on the basic environmental characteristics, the microclimate zone within this collection area is consistent with the collection area. The detection area is divided according to soil type, and the sub-identifier is XZ-001.

[0028] The soil in the sampling area was divided into three layers according to vertical depth: the first layer was 0-20 cm, the second layer was 20-40 cm, and the third layer was 40-60 cm. Environmental sensors were deployed in the detection area, including a humidity sensor for measuring atmospheric humidity, a temperature sensor for measuring atmospheric temperature, a light sensor for measuring light intensity, a cup anemometer for measuring wind speed, and a soil moisture meter buried in each layer, with the soil moisture meter buried in the middle of each soil layer.

[0029] By continuously monitoring environmental parameters without flower planting using sensors and combining this with stratified soil volumetric water content data, the weight coefficients of the baseline correlation model for each layer were obtained through data fitting. The first-level baseline correlation model is S. 0，1 =0.5H-0.2T-0.001L-0.1V, where a=0.5, b=-0.2, c=-0.001, d=-0.1; The second layer is S. 0，2 =0.45H-0.22T-0.0009L-0.12V, where a=0.45, b=-0.22, c=-0.0009, d=-0.12; The third layer is S 0，3= 0.4H-0.25T-0.0008L-0.15V, where a=0.4, b=-0.25, c=-0.0008, d=-0.15.

[0030] After associating the area identifier HZ-001, sub-identifier XZ-001, environmental parameters, and baseline association models at each layer, upload them to the cloud.

[0031] The planting area is a flower bed within the detection area, with clear boundaries and a constant volume, planted entirely with floribunda roses. The area identifier HZ-001 and sub-identifier XZ-001 of the planting area are obtained, and the corresponding baseline association models for each layer are retrieved from the cloud.

[0032] Environmental sensors, identical to those used in step S1, were deployed within the planting area. On May 1, 20XX, the environmental parameters measured were: atmospheric humidity 70%, atmospheric temperature 25℃, light intensity 8000 lux, and wind speed 2 m / s. Soil volumetric water content was obtained layer by layer using the sensors when flowers were present. Weighting coefficients for the flower influence model at each layer were then fitted. The first-level flower influence model is S. 1，1 =0.45H-0.18T-0.0009L-0.08V, where e=0.45, f=-0.18, g=-0.0009, h=-0.08; The second layer is S. 1，2=0.4H-0.2T-0.0008L-0.1V, where e=0.4, f=-0.2, g=-0.0008, h=-0.1; The third layer is S 1，3 =0.35H-0.22T-0.0007L-0.12V, where e=0.35, f=-0.22, g=-0.0007, h=-0.12.

[0033] Based on the baseline correlation model and the flower impact model, the unit water consumption of flowers in each layer is calculated: First layer ΔS1=S 0，1 -S 1，1 =21.8% - 19.64% = 2.16%; Following this logic, we obtain ΔS2 = 2.16% for the second layer and ΔS3 = 1.89% for the third layer. The planting area is 100 square meters, and each soil layer is 0.2 meters thick. Therefore, the volume of each soil layer is Vs = 100 × 0.2 = 20 cubic meters, and the soil bulk density is ρ = 1300 kg / m³. The actual water consumption of each layer is calculated as follows: First layer M1 = ΔS1 × V s ×ρ=2.16%×20×1300=561.6 kg; The second layer, M2, weighs 2.16% × 20 × 1300 = 561.6 kg. The third layer, M3, is calculated as follows: M3 = 1.89% × 20 × 1300 = 491.4 kg.

[0034] The timestamp corresponding to the actual water consumption of each floor is May 1, 20XX.

[0035] Cameras were installed in the planting area, with a preset acquisition frequency of one flower image taken at 10:00 AM daily. The image timestamps were recorded in relation to the water consumption timestamps. Image features were extracted using the convolutional layers of a convolutional neural network. The leaf texture feature quantization value for the floribunda rose was 85, the flower pixel ratio was 30%, and the flower coverage was 60%.

[0036] Based on the water consumption timestamps, a time series of actual water consumption for each floor is generated for 30 days from May 1st to May 30th, 20XX, with each floor's time series being independent. A sliding window step size of n=5 days is set, and the daily average rate of change in water consumption for each floor within the sliding window is calculated, with a preset threshold of 10%. First layer: Average daily water consumption was 620 kg from days 11 to 15, and 685 kg from days 12 to 16, with a change rate of approximately 10.48%. May 12th was marked as the first layer's specific change point. Average daily water consumption was 680 kg from days 21 to 25, and 580 kg from days 22 to 26, with a change rate of approximately -14.71%. May 22nd was marked as another first layer's specific change point. Second layer: Average daily water consumption was 610 kg from day 10 to 14, and 675 kg from day 11 to 15, with a change rate of approximately 10.66%. May 11th was marked as the specific change point for the second layer. Average daily water consumption was 670 kg from day 20 to 24, and 570 kg from day 21 to 25, with a change rate of approximately -14.93%. May 21st was marked as another specific change point for the second layer. The third layer: the average daily water consumption was 590 kg from day 12 to 16, and 650 kg from day 13 to 17, with a change rate of approximately 10.17%. May 13th was marked as the specific change point for the third layer. The average daily water consumption was 650 kg from day 22 to 26, and 550 kg from day 23 to 27, with a change rate of approximately -15.38%. May 23rd was marked as another specific change point for the third layer.

[0037] For 30 days of images of floribunda roses, K-means clustering was performed based on extracted varietal characteristics to form floribunda rose-specific clusters. Each cluster uniquely corresponds to a floribunda rose variety and is used for subsequent variety matching. Simultaneously, based on the specific change points in each layer, the water consumption time series for each layer was divided into specific water consumption segments. The first dedicated water consumption period is from May 1st to May 11th, May 12th to May 21st, and May 22nd to May 30th. The second dedicated water consumption period is from May 1st to May 10th, May 11th to May 20th, and May 21st to May 30th. The third exclusive water consumption period is from May 1st to May 12th, May 13th to May 22nd, and May 23rd to May 30th.

[0038] Each layer's dedicated water consumption segment is associated with a dedicated cluster of floribunda roses and stored, without any warnings.

[0039] Set T=2 days, divide each floor into a stable water consumption period and a transition period based on the specific change point of each floor, and calculate the specific water consumption coefficient for each floor: First layer: Stable water consumption segment A (May 1st - May 11th) average daily water consumption M0 = 560 kg, stable water consumption segment B (May 12th - May 21st) average daily water consumption M1 = 650 kg, dedicated stable water consumption coefficient K1 = 1.0 × 650 / 560 ≈ 1.16; transition segment (May 10th - May 13th) dedicated dynamic water consumption coefficient is calculated by linear interpolation; Second layer: The average daily water consumption of stable water consumption segment A (May 1st - May 10th) is M0 = 550 kg, and the average daily water consumption of stable water consumption segment B (May 11th - May 20th) is M1 = 640 kg. The dedicated stable water consumption coefficient K2 = 1.0 × 640 / 550 ≈ 1.17; the dedicated dynamic water consumption coefficient of the transition segment (May 9th - May 12th) is calculated by linear interpolation. The third layer: the average daily water consumption of stable water consumption segment A (May 1-May 12) is M0=530 kg, the average daily water consumption of stable water consumption segment B (May 13-May 22) is M1=620 kg, and the dedicated stable water consumption coefficient K3=1.0×620 / 530≈1.17; the dedicated dynamic water consumption coefficient of the transition segment (May 11-May 14) is calculated by linear interpolation.

[0040] Based on floribunda rose varieties, and combined with the specific water consumption coefficient for each layer, a specific predicted water consumption model for each layer of floribunda roses is constructed. This model is then stored in the cloud after being associated with the specific cluster of floribunda roses, the region identifier HZ-001, and the sub-identifier XZ-001. First-layer dedicated water consumption prediction model: M y，1 =ΔS1×V s ×ρ×K1=2.16%×20×1300×K1=561.6×K1; Second-layer dedicated water consumption prediction model: M y，2= ΔS2×V s ×ρ×K2=2.16%×20×1300×K2=561.6×K2; Third-layer dedicated water consumption prediction model: M y，3 =ΔS3×V s ×ρ×K3=1.89%×20×1300×K3=491.4×K3; Note: K1, K2, and K3 represent the dedicated stable or dynamic water consumption coefficients for the first, second, and third layers, respectively; The planting area to be predicted remains the same plot of land, and its area identifier HZ-001 and sub-identifier XZ-001 are obtained. At 10:00 AM on May 15, 20XX, actual images of the floribunda roses in the area to be predicted are captured by a camera. The actual images are then compared with the clusters stored in the cloud for similarity. Since the features of the actual image perfectly match the floribunda rose-specific cluster, the flower variety is confirmed as a floribunda rose. Based on the variety, area, and sub-identifier, the first, second, and third-level specific water consumption prediction models corresponding to the floribunda rose are retrieved from the cloud.

[0041] On May 15th, the environmental parameters measured were: atmospheric humidity 72%, atmospheric temperature 26℃, light intensity 8200 lux, and wind speed 2.3 m / s. Based on the specific water consumption prediction models for each floor and the current specific stable water consumption range, the specific predicted water consumption for each floor was calculated as follows: First layer: Current dedicated stable water consumption coefficient K1 = 1.16, M y，1 =2.214%×20×1300×1.16≈667.74 kg; Second layer: Current dedicated stable water consumption coefficient K2 = 1.17. M y，2 =2.251%×20×1300×1.17≈685.35 kg; Third layer: Current dedicated stable water consumption coefficient K3 = 1.17. M y，3 =1.931%×20×1300×1.17≈588.23 kg.

[0042] The preset irrigation time is 3 pm every day. The three layers of soil are irrigated in alternating layers: first, the first layer of soil is irrigated according to the predicted water consumption of the first layer. After an interval of 1 hour, the second layer of soil is irrigated according to the predicted water consumption of the second layer. After an interval of 1 hour, the third layer of soil is irrigated according to the predicted water consumption of the third layer, thus completing the soil environmental quality adjustment for the planting of floribunda roses on that day.

[0043] Example 2: Figure 2 As shown, this invention provides a soil environmental quality regulation system for flower cultivation. The system includes an environmental modeling module, a water consumption analysis module, a model training module, and an irrigation execution module. The environmental modeling module is used to construct a baseline correlation model for each layer. The water consumption analysis module is used to calculate the actual water consumption of flowers in each layer within the planting area. The model training module is used to construct a predicted water consumption model for each layer. The irrigation execution module is used to obtain the predicted water consumption of flowers in each layer and perform layered alternating irrigation. The output of the environmental modeling module is connected to the input of the water consumption analysis module; the output of the water consumption analysis module is connected to the input of the model training module; and the output of the model training module is connected to the input of the irrigation execution module.

[0044] The environmental modeling module also includes a partitioning unit and a construction unit; the partitioning unit is used to divide the collection area and the detection area according to the climate characteristics and terrain characteristics, and to obtain basic environmental characteristics; the construction unit is used to deploy environmental sensors and to construct a baseline correlation model for each layer based on environmental parameters.

[0045] The water consumption analysis module also includes a calling unit and a calculation unit; the calling unit is used to call the baseline correlation model of each layer and obtain the environmental parameters within the planting area; the calculation unit is used to calculate the unit water consumption of flowers in each layer and the actual water consumption of flowers in each layer.

[0046] The model training module also includes a segmentation unit and a coefficient unit; the segmentation unit is used to capture images of flowers through a camera, extract flower features, and divide the water consumption time series of each layer into a stable water consumption segment and a transition segment; the coefficient unit is used to calculate the stable water consumption coefficient and the dynamic water consumption coefficient, and construct the predicted water consumption model for each layer.

[0047] The irrigation execution module further includes a matching unit and an execution unit; the matching unit is used to compare the similarity between the actual image captured by the camera and the image in the cluster and obtain the matching image; the execution unit is used to calculate the predicted water consumption of each layer according to environmental parameters and perform layered alternating irrigation.

[0048] Finally, it should be noted that the above descriptions are merely preferred embodiments of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for regulating soil environmental quality for flower cultivation, characterized in that: The method includes the following steps: Step S1: Divide the collection area and obtain the basic environmental characteristics within the collection area; divide the soil in the collection area into several layers, deploy environmental sensors, obtain environmental parameters in the collection area, and obtain the soil volumetric water content without flowers in each layer, and construct a baseline correlation model for each layer. Step S2: Obtain the planting area of ​​the flowers and call the baseline association model for each layer; obtain the environmental parameters in the planting area through environmental sensors and obtain the soil volume water content when there are flowers in each layer; obtain the actual water consumption of the flowers in each layer according to the baseline association model. Step S3: Divide the actual water consumption of flowers in each layer into stable water consumption section and transition section, calculate the stable water consumption coefficient of the stable water consumption section and the dynamic water consumption coefficient of the transition section, and construct the predicted water consumption model for each layer. Step S4: Obtain the varieties of flowers in the planting area to be predicted, output the predicted water consumption of each layer of flowers according to the predicted water consumption model of each layer, and irrigate the flowers in layers alternately.

2. The method for regulating soil environmental quality for flower cultivation according to claim 1, characterized in that: Step S1 includes: Step S1-1: Divide areas with similar climate and topography into the same data collection area, and assign a unique area identifier to each data collection area; the basic environmental characteristics within the data collection area include climate characteristics, topographic characteristics, and environmental characteristics; The climate features include the annual average atmospheric humidity and the annual average atmospheric temperature range; the topographic features include altitude, slope and soil type; the environmental features include the annual average sunshine duration and wind speed. Step S1-2: Divide the microclimate zone within the collection area according to the basic environmental characteristics. Within each microclimate zone of the collection area, divide the detection area according to soil type and assign a sub-identifier to each detection area. The soil within the collection area is divided into several layers according to different vertical depths. Environmental sensors are deployed within the detection area, including humidity sensors, temperature sensors, light sensors, cup anemometers, and soil moisture meters. The environmental parameters include atmospheric humidity, atmospheric temperature, light intensity, and wind speed. The soil moisture meters are buried in layers. Steps S1-3: Construct a baseline correlation model for each layer of the detection area based on environmental parameters: S 0，j =aH+bT+cL+dV, where S represents the soil volumetric water content, S 0，j The soil volumetric water content of layer j without flowers is represented by H, atmospheric humidity by T, atmospheric temperature by L, light intensity by V, and wind speed by V. a, b, c, and d are weighting coefficients. The region identifier, sub-identifier, and environmental parameters are associated with the baseline association model of each layer and uploaded to the cloud.

3. The method for regulating soil environmental quality for flower cultivation according to claim 1, characterized in that: Step S2 includes: The planting area is a soil environment with physical boundaries and a constant volume within the boundaries; different flowers in the planting area are of the same variety; the area identifier and sub-identifier of the planting area are obtained, and the baseline association model associated with the cloud is called. Environmental sensors were deployed within the planting area to acquire environmental parameters, and the soil volumetric water content at each layer when flowers were present was obtained. A flower influence model was constructed for each layer according to the flower variety: S 1，j =eH+fT+gL+hV, where S represents the soil volumetric water content when flowers are present. 1，j denoted by , where represents the volumetric water content of the soil in the j-th layer when there are flowers, H represents atmospheric humidity, T represents atmospheric temperature, L represents light intensity, V represents wind speed, and e, f, g, and h are weighting coefficients, respectively. Based on the baseline correlation model and the flower impact model, the unit water consumption of flowers at each layer under the same environmental parameters is calculated as: ΔS j =S 0，j -S 1，j =(ae)H+(bf)T+(cg)L+(dh)V, ΔS j The water consumption per unit of flowers in the j-th layer is expressed as a percentage; the soil volume of the planting area is obtained, and the actual water consumption of flowers in different layers is calculated based on the soil volume: M j =ΔS j ×V s ×ρ, where M represents the actual water consumption of the flowers, M j ΔS represents the actual water consumption of the flowers in the j-th layer. j V represents the water consumption per unit of flowers in the j-th layer. s ρ represents the soil volume evenly distributed according to the number of layers within the planting area, and ρ represents the soil bulk density of the planting area; and records the timestamp of water consumption corresponding to the actual water consumption of flowers in each layer.

4. The method for regulating soil environmental quality for flower cultivation according to claim 1, characterized in that: Step S3 includes: Step S3-1: Set up a camera in the planting area, capture images of the flowers at a preset acquisition frequency, and record the image timestamps; Step S3-2: Extract flower varieties and flower features from the image through the convolutional layer of the convolutional neural network. The flower features include leaf texture features, the pixel ratio of flowers in the flower image, and flower coverage. The leaf texture features are the quantified values ​​of leaf texture complexity. The pixel ratio of flowers in the flower image is the ratio of the number of flower pixels to the total number of flower pixels in the image. The flower coverage is the percentage of the vertical projection area of ​​the flowers on the ground to the surface area of ​​the planting area. Step S3-3: Form a water consumption time series of the actual water consumption of flowers in each layer according to the order of water consumption timestamps; set a sliding window with a step size of n days, where n is an odd number; calculate the daily average change rate of the actual water consumption of flowers within the sliding window; when the daily average change rate of water consumption exceeds a preset threshold, mark the date of the (n-1) / 2th day from the start date of the sliding window as the change point; The layer closest to the soil surface is designated as the first layer. The water consumption time series of the first layer is divided into m segments according to the order of water consumption timestamps. The captured images are clustered using K-means clustering, and the images are clustered into m clusters according to the order of image timestamps. If the timestamps of the m segments do not match the timestamps of the corresponding m clusters within a preset timestamp range, an early warning is issued. If they match, the water consumption time series of each layer is divided into several water consumption segments based on the change points. Step S3-4: Divide the T days before and after the change point into transition segments. The T days before the transition segment belong to the stable water consumption segment A, and the T days after the transition segment belong to the stable water consumption segment B. And so on, divide the water consumption segments according to the stable water consumption segment and the transition segment. Calculate the average daily water consumption within each stable water consumption segment, and take the stable water consumption segment with the lowest average daily water consumption as the baseline segment. Define the stable water consumption coefficient K0 = 1.0 for the baseline segment; calculate the stable water consumption coefficient for each stable water consumption segment: Where K i M represents the steady water consumption coefficient of the i-th steady water consumption segment. i M0 represents the average daily water consumption of the i-th stable water consumption segment, and `M0 represents the average daily water consumption of the baseline segment. ； The dynamic water consumption coefficient of the transition section per day was calculated using linear interpolation. ； Where K(t) represents the dynamic water consumption coefficient on day t of the transition period, K A K represents the steady water consumption coefficient of steady water consumption section A. B The stable water consumption coefficient represents the stable water consumption section B; Step S3-5: Based on the stable water consumption coefficient and the dynamic water consumption coefficient, construct predictive water consumption models for different layers using flower varieties as a benchmark: M y，j =ΔS j ×V s ×ρ×K, where M y，j ΔS represents the predicted water consumption of flowers in the j-th layer; j V represents the water consumption per unit of flowers in the j-th layer; s ρ represents the soil volume evenly distributed according to the number of layers within the planting area; K represents the soil bulk density of the planting area; K represents the stable water consumption coefficient or dynamic water consumption coefficient of the j-th layer corresponding to the prediction date; the prediction water consumption models of different layers are stored in the cloud according to the region identifier, sub-identifier, cluster and flower variety.

5. The method for regulating soil environmental quality for flower cultivation according to claim 1, characterized in that: Step S4 includes: Step S4-1: Obtain the region identifier and sub-identifier of the planting area to be predicted, and retrieve the corresponding prediction water consumption model of different layers according to the flower variety; take actual images of flowers in the planting area to be predicted by camera, compare the similarity of the actual images with the images in the corresponding clusters, and take the image in the cluster with the highest similarity as the matching image. Step S4-2: Obtain the water consumption segment of each layer matched by the matching image, and obtain the stable water consumption segment or transition segment to which the matching image belongs based on the position of the water consumption segment in the water consumption time series of the corresponding layer; obtain the environmental parameters of the planting area to be predicted, obtain the unit water consumption of flowers in different layers based on the environmental parameters, calculate the predicted water consumption of the flower variety based on the predicted water consumption model, and perform layered alternating irrigation for layers where the predicted water consumption of flowers is not zero according to the preset time.

6. A soil environmental quality regulation system for flower cultivation, applied to the soil environmental quality regulation method for flower cultivation as described in any one of claims 1-5, characterized in that: The system includes an environmental modeling module, a water consumption analysis module, a model training module, and an irrigation execution module. The environmental modeling module is used to construct a baseline correlation model for each layer. The water consumption analysis module is used to calculate the actual water consumption of flowers in each layer within the planting area. The model training module is used to construct a predicted water consumption model for each layer. The irrigation execution module is used to obtain the predicted water consumption of flowers in each layer and perform stratified alternating irrigation. The output of the environmental modeling module is connected to the input of the water consumption analysis module; the output of the water consumption analysis module is connected to the input of the model training module; and the output of the model training module is connected to the input of the irrigation execution module.

7. A soil environmental quality regulation system for flower cultivation according to claim 6, characterized in that: The environmental modeling module also includes a partitioning unit and a construction unit; the partitioning unit is used to divide the collection area and the detection area according to climate characteristics and terrain characteristics, and to obtain basic environmental characteristics; The building unit is used to deploy environmental sensors and build a baseline correlation model for each layer based on environmental parameters.

8. A soil environmental quality regulation system for flower cultivation according to claim 6, characterized in that: The water consumption analysis module also includes a calling unit and a calculation unit; the calling unit is used to call the baseline correlation model of each layer and obtain the environmental parameters within the planting area; the calculation unit is used to calculate the unit water consumption of flowers in each layer and the actual water consumption of flowers in each layer.

9. A soil environmental quality regulation system for flower cultivation according to claim 6, characterized in that: The model training module also includes a segmentation unit and a coefficient unit; the segmentation unit is used to capture images of flowers through a camera, extract flower features, and divide the water consumption time series of each layer into a stable water consumption segment and a transition segment; the coefficient unit is used to calculate the stable water consumption coefficient and the dynamic water consumption coefficient, and construct the predicted water consumption model for each layer.

10. A soil environmental quality regulation system for flower cultivation according to claim 6, characterized in that: The irrigation execution module further includes a matching unit and an execution unit; the matching unit is used to compare the similarity between the actual image captured by the camera and the image in the cluster and obtain the matching image; the execution unit is used to calculate the predicted water consumption of each layer according to environmental parameters and perform layered alternating irrigation.

Images