A photovoltaic panel canopy M-type support rainwater drainage intelligent control method and system

Abstract

The application discloses a photovoltaic panel shed roof M-shaped support rainwater drainage intelligent control method and system, relates to the technical field of intelligent control, and comprises the following steps: acquiring a photovoltaic panel arrangement form, deploying a multilayer M-shaped support structure containing tubular structural members and transverse grooves, and connecting layers by using elastic pipelines; defining a tensor encoding library, deploying a first threshold at a flow guide system interface, and generating a conditional code element in combination with a multi-source environment sensor; activating a fluid calculation engine according to the conditional code element, initializing a micro-flow channel network, determining a support control strategy through conventional flow guiding and directional optimization, and regulating and controlling the M-shaped support structure. The application solves the technical problem that the existing control and regulation system function units are difficult to adapt to the rainwater drainage demand of the photovoltaic panel shed roof M-shaped support, cannot be accurately and dynamically regulated and controlled, and result in poor rainwater drainage effect, and achieves the technical effects of accurately adapting to the rainwater drainage of the photovoltaic panel shed roof M-shaped support, realizing dynamic regulation and control, and improving the rainwater drainage effect.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of intelligent control, and in particular to a photovoltaic panel shed roof M-type support rainwater drainage intelligent control method and system. BACKGROUND

[0002] The photovoltaic panel shed roof M-type support rainwater drainage is of great significance to ensure the stability of the photovoltaic system and improve the rainwater utilization efficiency, and intelligent control is the key to realizing efficient drainage. In the prior art, for the photovoltaic panel shed roof M-type support rainwater drainage, conventional control means are often used, combined with simple sensing and fixed strategy regulation and control. However, the M-type support structure and rainwater flow are complex, and the traditional control mode is difficult to accurately adapt to the multi-layer architecture of the support and the characteristics of the micro-flow network, and cannot dynamically optimize the flow guiding strategy according to the real-time environment, which easily causes the rainwater drainage to be not timely and not uniform, affecting the operation of the photovoltaic panel and the rainwater recovery effect, and is difficult to meet the efficient and intelligent drainage requirements. SUMMARY

[0003] The present application provides a photovoltaic panel shed roof M-type support rainwater drainage intelligent control method and system, which is used to solve the technical problem that the existing control and regulation system function units are difficult to adapt to the photovoltaic panel shed roof M-type support rainwater drainage requirements, cannot accurately dynamically regulate and control, and result in poor rainwater drainage effect.

[0004] In a first aspect, the present application provides a photovoltaic panel shed roof M-type support rainwater drainage intelligent control method, which comprises: obtaining the arrangement form of the photovoltaic panel and deploying the M-type support structure, wherein the M-type support structure is a multi-layer architecture, and the levels include tubular structural members, horizontal grooves, vertical grooves and auxiliary components, and the layers are connected based on elastic pipes; by defining a tensor encoding library, deploying a first threshold at the data interface of the flow guiding system, returning with multi-source environmental sensing, performing environmental tensor decomposition and flow guiding element conversion, and generating conditional code elements, wherein the conditional code elements are element freedom constraints; according to the conditional code elements, activating the fluid calculation engine deployed by the flow guiding system and initializing the built-in micro-flow network, to determine the support control strategy by first regular flow guiding and second directional synapse optimization, and regulating and controlling the M-type support structure, wherein the micro-flow network is deployed based on the M-type support structure.

[0005] In a second aspect of the present application, a photovoltaic panel shed M-type support rainwater drainage intelligent control system is provided, the system comprising: an M-type support structure deployment module for obtaining the arrangement form of the photovoltaic panel and deploying an M-type support structure, wherein the M-type support structure is a multi-layer architecture, the levels including tubular structural members, horizontal grooves, vertical grooves and auxiliary components, and the layers are connected based on elastic pipes; a conditional symbol acquisition module for defining a tensor encoding library, deploying a first threshold at the data interface of the diversion system, returning with multi-source environmental sensing, performing environmental tensor decomposition and diversion element conversion, and generating a conditional symbol, wherein the conditional symbol is an element freedom constraint; and an M-type support structure regulation and control module for activating the fluid computing engine deployed by the diversion system and initializing the built-in microfluid channel network according to the conditional symbol, determining the support control strategy through first regular diversion and second directional synapse optimization, and regulating and controlling the M-type support structure, wherein the microfluid channel network is deployed based on the M-type support structure.

[0006] The one or more technical solutions provided in the present application have at least the following technical effects or advantages:

[0007] In the present application, the photovoltaic panel shed M-type support multi-layer architecture is deployed, the tensor encoding library is constructed to process multi-source environmental sensing data to generate a conditional symbol, the fluid computing engine containing a double channel is activated, the control strategy is determined through parallel decision and interaction, the sub-strategy set is released under the spatiotemporal code constraint, the intelligent regulation and control of the M-type support rainwater drainage is realized, the photovoltaic panel shed rainwater drainage is more efficient and accurate, the photovoltaic facility rainwater management needs are adapted, the technical effect of accurately adapting the photovoltaic panel shed M-type support rainwater drainage is achieved, dynamic regulation and control is realized, and the rainwater drainage effect is improved. BRIEF DESCRIPTION OF DRAWINGS

[0008] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings needed in the embodiment description will be briefly introduced. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor.

[0009] Figure 1 is a flow diagram of a photovoltaic panel shed M-type support rainwater drainage intelligent control method provided by an embodiment of the present application.

[0010] Figure 2 is a structural diagram of a photovoltaic panel shed M-type support rainwater drainage intelligent control system provided by an embodiment of the present application.

[0011] Explanation of reference signs: M-type support structure deployment module 1, conditional symbol acquisition module 2, M-type support structure regulation and control module 3. DETAILED DESCRIPTION

[0012] The application provides a photovoltaic panel shed roof M-shaped support rainwater drainage intelligent control method and system, which is used for solving the technical problem that the existing control and adjustment system function unit is difficult to adapt to the rainwater drainage demand of the photovoltaic panel shed roof M-shaped support, cannot be accurately and dynamically regulated, and leads to poor rainwater drainage effect.

[0013] The technical solutions in the embodiments of the application will be clearly and completely described in the specification of the application combined with the drawings in the embodiments of the application. Obviously, the described embodiments are only part of the embodiments of the application, not all the embodiments. Based on the embodiments in the application, all other embodiments obtained by those skilled in the art without creative labor belong to the protection scope of the application.

[0014] It should be noted that the terms "first", "second" and the like in the specification and the above drawings of the application are used to distinguish similar objects, and do not necessarily indicate a specific order or sequence. It should be understood that the data thus used can be interchanged under appropriate circumstances, so that the embodiments of the application described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "include" and "have" and any variations thereof are intended to cover non-exclusive inclusion, for example, a process, method, system, product or server including a series of steps or units does not have to be limited to those steps or units clearly listed, but can include other steps or modules not clearly listed or inherent to these processes, methods, products or devices.

[0015] Embodiment one, as shown in a photovoltaic panel shed roof M-shaped support rainwater drainage intelligent control method, wherein the method comprises: Figure 1

[0016] Step A100: Obtain the arrangement form of the photovoltaic panel, and deploy the M-shaped support structure, wherein the M-shaped support structure is a multi-layer architecture, the hierarchy includes tubular structural members, horizontal grooves, vertical grooves and auxiliary components, and the layers are connected based on elastic pipes.

[0017] Specifically, first, the arrangement form of the photovoltaic panel is obtained, which is usually extracted by a person skilled in the art through field surveying or design drawings to obtain key parameters such as the horizontal spacing of the photovoltaic panel, the longitudinal arrangement density, the inclination angle, etc. Assuming that the horizontal spacing of the photovoltaic panel is 1 meter, one column is arranged every 5 meters in the longitudinal direction, and the inclination angle is 15 degrees, these data will directly determine the size and layout of the M-shaped support.

[0018] ​Next, based on the obtained arrangement parameters, the M-shaped support structure is deployed, and the structural components are shown in Table 1. The structure is a multi-layer architecture. The first layer deploys a tubular structural member, the diameter of which is determined according to the load-bearing calculation of the photovoltaic panel, assuming 8 cm, and the lateral spacing is consistent with the photovoltaic panel, i.e. 1 m, to ensure stable support for the photovoltaic panel. Above the tubular structural member, a transverse slot is deployed at the gap position corresponding to the photovoltaic panel, with a slot width of 10 cm to adapt to the possible water flow path; the longitudinal slot is perpendicular to the transverse slot and has a spacing corresponding to the 5 m interval of the longitudinal arrangement of the photovoltaic panel, forming a basic hierarchical flow guide structure.

[0019] The deployment of auxiliary components is also determined according to the arrangement of the photovoltaic panel and the support hierarchy. The micro vortex generator is installed at the intersection of the transverse slot and the longitudinal slot, with one installed every 20 cm, and the elastic clamping connector is used to connect the hierarchical structure, using a material with an elastic coefficient of 5 N / mm to provide a certain deformation buffering capacity when the interlayer is connected by an elastic pipe. The length of the elastic pipe is designed to be 35 cm according to the 30 cm spacing between the layers, ensuring that the interlayer connectivity is maintained even when the support is slightly deformed due to environmental changes.

[0020] By obtaining the arrangement form of the photovoltaic panel and deploying the M-shaped support multi-layer architecture including the tubular structural member, the transverse slot, the longitudinal slot, and the auxiliary components according to the specific parameters, and connecting the layers with elastic pipes, the precise adaptation of the support structure to the photovoltaic panel is achieved, providing a stable physical basis for subsequent support control analysis.

[0021] Table 1: M-shaped support structure composition table

[0022]

[0023] Step A200: By defining a tensor encoding library, a first threshold is deployed at the data interface of the flow guide system, and environmental tensor decomposition and flow guide element conversion are performed on the multi-source environmental sensing return, to generate a conditional code element, wherein the conditional code element is a constraint on the freedom of elements.

[0024] In the embodiments of the present application, the flow guide system is a system for realizing intelligent control of rainwater drainage of the M-shaped support of the photovoltaic panel canopy. The first threshold is a threshold deployed at the data interface of the flow guide system, which is used to filter the data returned by the multi-source environmental sensor, to ensure that the environmental data entering the subsequent processing is effective. The conditional code element is generated by performing environmental tensor decomposition and flow guide element conversion on the environmental data filtered by the first threshold, and is a constraint on the freedom of elements, which integrates the adjustable element constraints corresponding to the environmental data.

[0025] Optionally, the definition of the tensor encoding library comprises: constructing a first tensor matrix with the correlation degree as the weight value through tensor decomposition and weight filtering for each environmental dimension, defining a position-coding adjustable element for the M-type support assembly architecture to determine a second guide element matrix, and then determining the tensor encoding library based on the above. The specific steps are described in detail in A210-A230.

[0026] After the definition of the tensor encoding library is completed, a first threshold is then deployed at the data interface of the guide system. The threshold is set based on the typical parameter range of the environmental dimensions in the tensor encoding library, for example, the rainfall intensity threshold is set to 0-50 mm / h, the wind speed threshold is set to 0-10 m / s, the dust particle size threshold is set to 0-100 μm, and the pulse coefficient fluctuation amplitude threshold is set to 0-50%. When the multi-source environmental sensor collects data, the real-time returned information such as rainfall intensity, wind speed, dust particle size, and pulse coefficient needs to be first screened by the first threshold to eliminate abnormal data that exceeds the set range, such as instantaneous rainfall intensity of 55 mm / h or wind speed of 12 m / s, which will be filtered to ensure that the environmental data entering the subsequent processing meets the valid range.

[0027] Subsequently, environmental tensor decomposition is performed on the environmental data that passes through the first threshold. Taking the returned comprehensive environmental data as an example, which contains rainfall intensity of 35 mm / h, wind speed of 7 m / s, dust particle size of 60 μm, and pulse coefficient of 25%, based on the division of the environmental dimensions of the tensor encoding library, the comprehensive data is decomposed into rainfall sub-tensor, wind environment sub-tensor, dust sub-tensor, and pulse sub-tensor. Each sub-tensor corresponds to a single environmental dimension, for example, the rainfall sub-tensor only contains rainfall intensity information of 35 mm / h, the wind environment sub-tensor only contains wind speed information of 7 m / s, and the structure of each sub-tensor is consistent with the environmental dimension structure in the first tensor matrix.

[0028] Then, guide element conversion is performed. The decomposed environmental sub-tensors are matched with the correlation in the tensor encoding library to convert into corresponding adjustable element constraints. For example, the rainfall sub-tensor 35 mm / h corresponds to the 30-50 mm / h interval in the first tensor matrix with a correlation degree of 0.8, and according to the mapping of the pipe structure inclination in the encoding library, it is converted into an adjustment constraint of inclination 20-30°; the wind environment sub-tensor 7 m / s corresponds to the 5-10 m / s interval, matches the mapping relationship of the horizontal slot flow direction, and is converted into a flow direction constraint of -15° to -5°; the dust sub-tensor 60 μm corresponds to the 50-100 μm interval, and is converted into a constraint of micro vortex generator rotational flow level 3-5.

[0029] Finally, all the converted adjustable element constraints are integrated to generate conditional symbols. For example, the rainfall inclination constraint, wind flow direction constraint, and dust swirl constraint are integrated to form the symbol T20-30-H-15--5-V3-5, where each part corresponds to the inclination of the tubular structure, the flow direction of the transverse slot, and the swirl level of the micro vortex generator, respectively.

[0030] Through the above steps, the generated conditional symbols accurately associate the environmental data with the constraint range of the adjustable elements of the support, providing a clear and structured element freedom basis for the subsequent development of the support control strategy.

[0031] Step A300: According to the conditional symbols, activate the fluid computing engine deployed in the flow guiding system and initialize the built-in micro channel network to determine the support control strategy through first regular flow guiding and second directional synapse optimization, and regulate the M-type support structure, wherein the micro channel network is deployed based on the M-type support structure.

[0032] In the embodiments of the present application, the micro channel network is a twin flow field network of the M-type support structure, consistent with the network state of the M-type support structure, and is deployed based on the M-type support structure.

[0033] In an embodiment of the present application, when the fluid computing engine is activated and the micro channel network is initialized according to the conditional symbols, the conditional symbols need to be parsed and verified first. The conditional symbols contain element freedom constraints, such as the inclination of the tubular structure 0-30°, the flow direction of the transverse slot -15° to 15°, and the swirl level of the micro vortex generator 1-5 levels, etc. By comparing the standard constraint range in the tensor encoding library, it is confirmed whether each parameter in the symbol is within the valid interval, for example, if the inclination constraint in a certain symbol is 35°, which is out of the range of 0-30°, it is determined as an invalid symbol and triggers the re-generation process to ensure that the symbol input to the fluid computing engine meets the specifications.

[0034] After verification, the fluid computing engine deployed in the flow guiding system is activated. The system will first check the current state of the engine, if it is in sleep mode, start the wake-up program, load the preset basic calculation model, including the flow field dynamics equation, component response parameters, etc., the response time of the whole activation process is controlled within 0.5 seconds. At the same time, call the latest back data of multi-source environmental sensing, such as real-time rainfall intensity 20mm / h, wind speed 3m / s, as the initial environmental parameters after the engine starts, to ensure that the engine obtains real-time environmental background information.

[0035] Subsequently, the built-in microfluid network is initialized. The microfluid network, as a twin flow field network of the M-shaped stent structure, needs to update the network parameters according to the actual state of the current stent, including the current inclination angle of the tubular structure, the real-time flow direction of the transverse groove, the bending degree of the elastic pipe, etc. Through the 200 groups of real-time parameters collected by the position sensor and the state monitoring module, the geometric model of the microfluid network, the flow channel resistance coefficient, the component connection relationship, etc. are calibrated, so that the deviation between the network state and the actual state of the stent is controlled within ±0.5° (angle parameter) and ±1mm (length parameter), and it is ensured that the initialized microfluid network can accurately reflect the current physical state of the stent.

[0036] After initialization, the system will perform a completeness check to check whether all nodes of the microfluid network respond normally and whether the fluid calculation engine is loaded successfully. After passing the verification, an ready signal is output to provide a basis for subsequent determination of the stent control strategy through the first regular conduction and the second directional synapse optimization.

[0037] Next, the stent control strategy is determined through the first regular conduction and the second directional synapse optimization, including classifying two types of code elements according to conditional code elements in regular and directional dimensions, respectively importing corresponding channels to perform parallel decision and intermediate feature lateral interaction, and outputting the strategy. The specific steps are described in detail in A340-A350.

[0038] Finally, the M-shaped stent structure is regulated, including identifying the stent control strategy, decomposing into a set of sub-strategies with position coding based on component driving, introducing time stamp and position code constrained space-time code, and distributing through the conduction system, executing component control management. The specific steps are described in detail in A360-A370.

[0039] By analyzing the verification conditional code elements, activating the fluid calculation engine and initializing the microfluid network, the accurate synchronization and effective start of the calculation engine and the stent twin network are ensured, and a reliable calculation environment and model basis are provided for the efficient generation of subsequent stent control strategies.

[0040] Further, the step A200 in the method provided by the embodiment of the present application comprises:

[0041] A210: Perform tensor decomposition and weight filtering for each environmental dimension to construct a first tensor matrix, wherein the weight is defined by the dredging correlation degree.

[0042] A220: Define adjustable elements according to the component architecture of the M-shaped stent structure to determine a second conduction element matrix, wherein the adjustable elements are identified with position coding.

[0043] A230: Determine the tensor code library according to the first tensor matrix and the second conduction element matrix.

[0044] In the embodiments of the present application, the environmental dimension refers to various environmental related dimensions for tensor decomposition and weight filtering, such as rainfall, wind direction and speed, dust, and other environmental factor dimensions affecting the M-type support dredging. The dredging correlation degree is an index for defining the weight in the first tensor matrix, reflecting the correlation degree between the environmental dimension factors and the M-type support rainwater dredging effect. The adjustable element is an adjustable element defined for the component architecture of the M-type support structure, and is identified with a position code, such as the adjustable parameters of the tubular structure, the transverse groove, the longitudinal groove, and the auxiliary components.

[0045] Specifically, in defining the tensor code library, first, each environmental dimension is processed. Taking rainfall intensity as an example, when performing tensor decomposition in each interval, the collected rainfall intensity data (example: 0-50 mm / h) is divided into 10 discrete intervals at an interval of 5 mm / h, such as 0-5 mm / h, 5-10 mm / h,..., 45-50 mm / h, each interval is taken as an independent dimension component, and an initial rainfall intensity tensor is constructed, with a dimension of 10 x n, n being the sample number.

[0046] Subsequently, the support dredging effect parameters corresponding to each interval in the historical operation data are called, including the average flow rate of rainwater in the microchannel network, such as 0.2 m / s for the 0-5 mm / h interval and 1.5 m / s for the 30-50 mm / h interval; the water accumulation area ratio of the support regions, such as 3% for the 0-5 mm / h interval and 25% for the 30-50 mm / h interval; and the drainage completion time, such as 15 minutes for the 0-5 mm / h interval and 5 minutes for the 30-50 mm / h interval.

[0047] Based on these data, the initial tensor is decomposed using a high-order singular value decomposition algorithm to extract the core feature vector of each interval, which reflects the typical characteristics of the rainfall intensity of the interval. Then, the correlation degree is calculated by the cosine similarity between the core feature vector and the effect parameter vector, i.e., the higher the similarity, the greater the correlation degree. For example, the similarity between the feature vector of the 30-50 mm / h interval and the effect vector of high flow rate, high water accumulation ratio, and short drainage time is 0.8, so the dredging correlation degree is set to 0.8; the similarity between the feature vector of the 0-10 mm / h interval and the effect vector of low flow rate and low water accumulation ratio is 0.2, so the dredging correlation degree is set to 0.2.

[0048] After the correlation degree is calculated, the weights are filtered to eliminate abnormal intervals with a correlation degree lower than 0.1, such as individual data points of extreme instantaneous rainfall, and the effective intervals and their correlation degrees are retained. Finally, these intervals, corresponding correlation degrees, and feature parameters are integrated to form a sub-matrix of the rainfall dimension.

[0049] For wind direction and wind speed, 0-10 m / s wind speed and 8 azimuth wind direction data are collected, which are decomposed into wind speed levels 0-2 m / s, 2-5 m / s, 5-10 m / s and wind direction intervals, the correlation degrees thereof with the channel blockage risk are calculated, and are integrated into a wind environment sub-matrix, while the dust particle size, 0-100 μm, is divided by 20 μm intervals. The pulse coefficient reflects the stability of rainfall, and the corresponding correlation degree is calculated. The calculation process is the same as the above-mentioned rainfall intensity step, which will not be described in detail due to the limited length of the specification. These sub-matrices are combined to construct the first tensor matrix.

[0050] Next, the adjustable elements of the M-type bracket assembly architecture are determined. For example, the inclination angle of the tubular structure is adjustable in the range of 0-30°, and each adjustment position corresponds to a position code T1 to T31; the flow direction adjustment angle of the transverse slot is -15° to 15°, divided by 1° intervals, coded H1 to H31; the rotational flow adjustment level of the micro vortex generator is divided into 1-5 levels, corresponding to different rotational flow intensities, coded V1 to V5; the elastic adjustment amount of the elastic clamping connector is 0-20 mm, set by 2 mm steps, coded E1 to E10. The parameter range, adjustment accuracy and position coding system of these adjustable elements are sorted to form a second flow guide element matrix, wherein each element is bound to a specific component of the bracket through position coding.

[0051] Finally, the tensor coding library is determined according to the first tensor matrix and the second flow guide element matrix, including introducing a unified coding mode for homologous and heterogeneous elements, cascading the two matrices and using the mode to code the matrices to determine the coding library, and the specific steps are described in detail in A231-A232.

[0052] Through the above steps, the tensor coding library constructed realizes the accurate mapping of environmental factors and adjustable elements of the bracket, and provides a unified and closely related coding basis for the generation of subsequent control strategies.

[0053] Further, the method provided in the embodiment of the application comprises the following steps:

[0054] A231: introducing a coding mode, wherein the coding mode is a unified coding mode for homologous and heterogeneous elements.

[0055] A232: by cascading the first tensor matrix and the second tensor matrix, using the coding mode, and by coding the matrices, the tensor coding library is determined.

[0056] Optionally, first, a unified coding mode is designed for isomeric elements. The mode provides that the code is composed of environmental dimension identification, guiding correlation value, adjustable element type, position code and parameter range, each part is represented by fixed number of characters or numbers, ensuring that one code corresponds to one objective dimension. For example, the environmental dimension identification uses 2-bit letters, such as R for rainfall and W for wind speed; the guiding correlation uses 1-bit decimal, such as 0.8; the adjustable element type uses 2-bit letters, such as T for the inclination of tubular structural member and H for the flow direction of transverse slot; the position code uses 3-bit numbers, such as 015 for the specific component position; and the parameter range uses 4-bit numbers, such as 2030 for 20-30°, so as to remove the interpretation redundancy in feature analysis and improve the coding uniformity.

[0057] Then, the first tensor matrix and the second guide element matrix are cascaded. The first tensor matrix contains environmental data such as rainfall intensity and wind speed, and the second guide element matrix contains adjustable element information such as the inclination of tubular structural member and the flow direction of transverse slot. When cascading, the environmental dimension and the corresponding adjustable element are paired according to the correlation relationship, for example, the rainfall intensity of 30-50 mm / h is paired with the inclination of tubular structural member of 20-30°, and the wind speed of 5-10 m / s is paired with the flow direction of transverse slot of -15° to -5°, to form a joint matrix containing environmental-element correlation relationship.

[0058] Finally, the cascaded matrix is coded using the above coding mode. For example, for the pairing of rainfall intensity of 30-50 mm / h, correlation degree of 0.8 and inclination of tubular structural member of 20-30°, and position code of 015-020, the code R0.8T0152030 is generated, where R represents rainfall, 0.8 is the correlation degree, T represents the inclination, and 015 and 2030 correspond to the starting value of the position code and the parameter range respectively; for the pairing of wind speed of 5-10 m / s, correlation degree of 0.7 and flow direction of transverse slot of -15° to -5°, and position code of 001-011, the code W0.7H001-15-5 is generated, and the coding conversion of all pairings is completed in a similar manner. Finally, the codes are integrated to determine the tensor coding library.

[0059] By introducing the unified coding mode, cascading the matrix and coding, the tensor coding library constructed realizes the efficient correlation between environmental factors and adjustable elements of the stent, eliminates the interpretation redundancy of isomeric elements, and provides a unified and accurate coding basis for the subsequent rapid call of control strategies.

[0060] Further, the step A300 in the method provided by the embodiment of the application comprises:

[0061] A310: The micro-channel network is a twin flow field network of the M-shaped stent structure, wherein the micro-channel network is consistent with the network state of the M-shaped stent structure.

[0062] A320: constructing a fluid computing engine according to the microfluidic network.

[0063] In the embodiments of the present application, the twin flow field network is a microfluidic network of an M-shaped stent structure, which is consistent with the network state of the M-shaped stent structure.

[0064] Specifically, before constructing the microfluidic network, the actual state data of the M-shaped stent structure after the last adjustment needs to be obtained. This includes the current inclination angle of the tubular structural member, the flow direction angle of the transverse groove, the depth of the longitudinal groove, the opening state of the micro vortex generator in the auxiliary assembly, and the extension amount of the elastic clamping connector, as well as the bending degree and the communication state of the interlayer elastic pipeline. These data are collected in real time by the position sensor, angle sensor and state monitoring module deployed on the stent, forming a stent state data set containing 200 parameters.

[0065] According to the collected stent state data, the corresponding microfluidic network is constructed. The network takes the physical structure of the M-shaped stent as the prototype, maps the tubular structural member into a main pipe with a diameter of 8 cm, maps the transverse groove and the longitudinal groove into a transverse flow channel with a width of 10 cm and a longitudinal flow channel with a depth of 5 cm, respectively, and maps the micro vortex generator of the auxiliary assembly into a vortex simulation unit in the flow channel. The elastic clamping connector and the elastic pipeline are mapped into a flexible flow channel segment that can be extended and retracted. For example, the 25° inclination angle of the stent tubular structural member corresponds to the 25° inclination angle of the main pipe in the microfluidic network, and the 10 mm extension amount of the elastic pipeline corresponds to the 10 mm length change of the flexible flow channel segment, ensuring that the geometric parameters and component states of the microfluidic network are completely matched with the M-shaped stent structure.

[0066] After the microfluidic network is constructed, the state synchronization verification is performed. By comparing the current key parameters of the stent with the corresponding parameters of the microfluidic network, such as the flow direction of the transverse flow channel, the deviation value is calculated, and when all parameter deviations are controlled within the range of ±0.5° angle and ±1 mm length / extension amount, it is determined that the states of the two are consistent. If there is a deviation beyond the range, such as a 2 mm deviation between the length of a certain flexible flow channel in the microfluidic network and the extension amount of the elastic pipeline of the stent, the microfluidic network parameters are corrected according to the actual state of the stent until the synchronization accuracy meets the standard.

[0067] On the basis of state synchronization, subsequent analysis is carried out based on the microfluidic network. For example, when the stent needs to be adjusted again due to environmental changes, the microfluidic network will be updated to the state after the adjustment, such as the inclination angle of the tubular structural member being adjusted to 30°, the inclination angle of the microfluidic main pipe is also updated to 30°, and the fluid flow is simulated based on this state to provide an accurate flow field model for the fluid computing engine.

[0068] Finally, the fluid computing engine is constructed according to the microfluid network, including the first channel for supervising and training the conditional symbol-based conventional scene decision optimization and the second channel for supervising and training the directional conditional trigger decision optimization, and the two channels are cascaded and lateral interaction is established to determine the engine, and the specific steps are described in detail in A321-A323.

[0069] By keeping the microfluid network consistent with the network state of the M-shaped support structure, it is ensured that the analysis based on the network can truly reflect the actual situation of the support structure, and a precise and synchronous model basis is provided for the construction of the subsequent fluid computing engine and the development of the control strategy.

[0070] Further, the step A320 in the method provided by the embodiment of the application comprises:

[0071] A321: The first channel is supervised and trained by embedding the microfluid network to guide the decision and optimization of the conventional scene based on the conditional symbol.

[0072] A322: The second channel is supervised and trained by embedding the microfluid network to guide the decision and optimization of the directional conditional trigger based on the conditional symbol.

[0073] A323: The first channel and the second channel are cascaded, and lateral interaction is established to determine the fluid computing engine.

[0074] Specifically, when constructing the fluid computing engine, the first channel is first supervised and trained. The microfluid network is embedded into the training model, and the conditional symbols of the conventional scene are selected as the training data. These symbols correspond to stable environments such as rainfall intensity 10-30 mm / h, wind speed 2-5 m / s, and dust particle size <50 μm, and contain adjustable element constraints such as pipe structure inclination 10-20° and transverse slot flow direction -5° to 5°. Genetic algorithm is used for optimization, and the minimum fluid resistance in the flow channel is used as the target. The decision parameters corresponding to each symbol, such as inclination adjustment step and slot body turning speed, are iteratively optimized. 50 candidate solutions are generated each iteration, and the top 20% of solutions with the best fitness are selected for crossover and mutation. After 100 iterations, the deviation of the decision result from the optimal solution is controlled to be within 3%. Through repeated training of 1000 groups of conventional scene data, until the decision accuracy of the model on new conventional symbols reaches 95%, the first channel training is completed.

[0075] Then the second channel is supervised to train. Also with the microfluid network, the conditional symbols triggered by directional conditions are selected, including rainfall intensity > 30 mm / h, wind speed > 5 m / s, dust particle size ≥ 50 μm, etc. Special environment, the symbol contains the start of the micro vortex generator, the expansion of the elastic clamping connector, etc. Trigger constraints. Reinforcement learning is used for optimization, and the drainage efficiency of the flow channel after triggering is used as a reward signal. When the vortex generator is turned on, the drainage speed is increased by ≥ 20%, and a positive reward is given, otherwise a penalty is given. For each group of trigger symbols, the model adjusts the vortex level (1-5 levels), the elastic expansion amount (5-20 mm) and other parameters through trial and error. After 500 groups of trigger scene training, the trigger response time is shortened to 0.5 seconds, and the decision effectiveness is more than 90%, the second channel training is completed.

[0076] After that, the first channel and the second channel are cascaded, and lateral interaction is established. The output layers of the two channels are connected to the same decision fusion module, the first channel processes continuous decision parameters in regular scenes, and the second channel processes trigger control signals in special scenes. Lateral interaction is achieved by sharing intermediate features, for example, the current flow channel pressure distribution data output by the first channel is transmitted to the second channel in real time, which helps it to determine the vortex intensity more accurately when triggered; The trigger state of the second channel is also fed back to the first channel, so that it avoids parameter conflicts in the triggered area in regular decision-making. For example, when the second channel triggers the vortex generator, the first channel will automatically adjust the inclination angle of the tubular structure in the corresponding area to avoid flow field disorder.

[0077] Through the above steps, the fluid calculation engine constructed can not only efficiently process support control decisions in regular environments, but also quickly respond to trigger requirements in special conditions, realizing the cooperation of regular and directional control, and providing a reliable calculation basis for accurate generation of support control strategies.

[0078] Further, the method provided in the embodiment of the application comprises the following steps:

[0079] A322-1: The auxiliary assembly at least includes a micro vortex generator and an elastic clamping connector.

[0080] A322-2: For the micro vortex generator, a first trigger condition is set, and for the elastic clamping connector, a second trigger condition is set.

[0081] A322-3: According to the first trigger condition and the second trigger condition, the second channel is constructed.

[0082] In the embodiments of the present application, the micro vortex generator is one of the auxiliary components of the M-shaped support structure, which triggers to meet the set first trigger condition, and is used to assist rainwater drainage in specific scenarios. The elastic clamping connector is one of the auxiliary components of the M-shaped support structure, which needs to meet the set second trigger condition, and can realize self-adaptive assembly between support levels through elastic deformation to adapt to different flow requirements.

[0083] Specifically, the micro vortex generator in the auxiliary component is mainly used for directional flow guiding, and adjusts the flow direction of the fluid in the flow channel by generating local vortex to avoid local water accumulation. The elastic clamping connector can realize self-adaptive assembly between support levels through elastic deformation, and adjust the connection tightness when the rain intensity fluctuates to adapt to different flow requirements. For example, when the rainfall is heavy, the upper flow channel overflows, and part of the water flows to the lower level. The elastic clamping connector can be stretched to 15-20 mm to expand the flow space. When the rainfall is small, it is contracted to 5-10 mm to maintain the sealing of the upper flow channel.

[0084] The first trigger condition is set for the micro vortex generator, and combined with the real-time monitoring data in the flow channel, when the rainfall continues for more than 10 minutes and exceeds 30 mm / h, or the fluid flow rate in a certain area of the flow channel is less than 0.3 m / s, which may form water accumulation, the trigger condition takes effect. At the same time, the trigger delay is set to not more than 2 seconds to ensure timely response. The second trigger condition is set for the elastic clamping connector, when the pressure sensor in the upper flow channel detects that the pressure exceeds 50 Pa, indicating that there is more water accumulation, or the displacement sensor monitors that the relative displacement between levels exceeds 8 mm, indicating that the stress is uneven, the trigger condition is activated, and the initial adjustment needs to be completed within 5 seconds after triggering.

[0085] When the second channel is constructed according to the first trigger condition and the second trigger condition, first, the parameter range and response threshold of the two types of trigger conditions are recorded into the decision system of the channel to establish the mapping relationship between the conditions and the actions of the auxiliary components. For example, when the first trigger condition is met, the system sends a start signal to the micro vortex generator to make it run at a 3-5 level rotational flow intensity; when the second trigger condition is met, a stretching instruction is sent to the elastic clamping connector to adjust to the corresponding length according to the measured pressure or displacement value. At the same time, the condition priority judgment is set in the channel, when the two types of conditions are met at the same time, the emergency condition of pressure exceeding 80 Pa or flow rate being less than 0.2 m / s is responded preferentially to ensure the processing efficiency in critical scenarios.

[0086] By clearly defining the functions of the auxiliary components, setting specific trigger conditions and constructing the second channel accordingly, the precise response to special flow guiding requirements under abnormal conditions is realized, so that the second channel can trigger the action of the auxiliary components in time under special conditions such as fluctuation of rain intensity and local water accumulation, and meet the demand of directional regulation.

[0087] Further, the step A300 in the method provided in the embodiments of the present application comprises:

[0088] A331: Identify the condition symbols, partition them with a preset distinction degree, and determine the partition state of the M-type support structure, wherein each partition corresponds to a group of partition condition symbols.

[0089] A332: If the number of partitions is greater than 1, perform calculation domain division and computing power allocation on the fluid calculation engine according to the partition state, and determine the calculation network, wherein the calculation network comprises a plurality of parallel computing nodes.

[0090] In one embodiment, after activating the fluid calculation engine, first, the condition symbols are deeply identified to extract the spatial features and environmental parameters contained therein. The condition symbols not only contain element degree of freedom constraints such as the inclination angle range of tubular structural members and the lateral groove flow direction angle, but also carry position encodings and associated environmental data (such as 001-200 identifying different regions), such as real-time wind power, orientation angle, etc. in each region. By analyzing the position information and environmental parameters in the symbols, a feature matrix is established, where each row corresponds to a position encoding, and the columns contain wind power, orientation, inclination angle constraints, etc. to provide data basis for subsequent partitioning.

[0091] Then, partitioning is performed with a preset distinction degree as the standard. The preset distinction degree is set as the difference threshold of environmental parameters and structural features, for example, when the wind power difference is > 2 m / s, the orientation angle difference is > 30°, and the inclination angle constraint range overlap degree is < 50%, it is determined as different partitions. Cluster analysis is performed on the 200 position encodings in the feature matrix. If the wind power in the east region 001-060 is 2-3 m / s, the orientation is 90° east; the wind power in the west region 061-120 is 4-6 m / s, the orientation is 270° west; the wind power in the south region 121-200 is 3-4 m / s, the orientation is 180° south, the wind power difference between the three is > 2 m / s, and the orientation difference is > 30°, then the three regions are divided into east, west, and south three partitions, each partition corresponds to a group of partition condition symbols, such as the east region symbols contain inclination angle 10-15°, the west region 15-20°, and the south region 12-18°.

[0092] After determining the partition state, the number of partitions is counted. If the number is greater than 1, such as the above three partitions, the calculation domain is divided according to the partition state. Each partition corresponds to an independent calculation domain, and the calculation domain boundary is consistent with the position encoding range of the partition, such as the east region calculation domain covers the micro-channel network corresponding to 001-060, and the computing power is allocated based on the partition complexity. The west region is allocated 40% of the computing power, corresponding to 2 computing cores, the east region and the south region are each allocated 30%, each with 1.5 cores, and the computing power is virtualized. Each calculation domain is deployed with a parallel computing node, and the nodes are connected through a high-speed bus to form a calculation network containing 3 nodes, and each node processes the fluid calculation task of the corresponding partition.

[0093] Through the above steps, the modular control of the large-coverage photovoltaic panel canopy is realized, the fluid calculation engine can accurately allocate resources according to the environmental and structural differences of different areas, ensure that the control decisions of each partition are generated efficiently, and adapt to the fine regulation and control requirements in complex scenarios.

[0094] Further, the step A300 in the method provided in the embodiments of the application includes:

[0095] A340: according to the conditional symbol, symbol dichotomization is performed in the conventional flow guiding dimension and the directional synapse dimension to determine the first conditional symbol and the second conditional symbol.

[0096] A350: the first conditional symbol is introduced into the first channel, the second conditional symbol is introduced into the second channel, parallel decision and intermediate feature lateral interaction are performed, and the support control strategy is output.

[0097] In the embodiments of the application, the conventional flow guiding dimension is one of the dimensions for dichotomizing the conditional symbol, corresponds to a conventional scene with stable environmental parameters, is used to generate the first conditional symbol, and is introduced into the first channel for conventional scene flow guiding decision and optimization. The directional synapse dimension is one of the dimensions for dichotomizing the conditional symbol, corresponds to a scene with large environmental parameter fluctuations or special processing requirements, is used to generate the second conditional symbol, and is introduced into the second channel for directional conditional trigger decision and optimization.

[0098] Optionally, when dichotomizing the conditional symbol, the division standard of the conventional flow guiding dimension and the directional synapse dimension is first determined. The conventional flow guiding dimension corresponds to a scene with stable environmental parameters, such as rainfall intensity of 5-15 mm / h and wind speed of 1-3 m / s. Such conditional symbols only include basic adjustment constraints such as pipe structure inclination of 5-15° and lateral groove flow direction of -5° to 5°. The directional synapse dimension corresponds to a scene with large environmental parameter fluctuations or special processing requirements, such as rainfall intensity > 30 mm / h and wind speed > 6 m / s. The symbol contains trigger type constraints such as micro vortex generator start and elastic clamping connector expansion. The conditional symbol is filtered by setting a threshold value. For example, when the rainfall intensity parameter in the symbol is > 20 mm / h or the wind speed is > 5 m / s, it is determined as the second conditional symbol, otherwise as the first conditional symbol.

[0099] After the first conditional symbol is introduced into the first channel, the built-in conventional decision model is called by the channel, the flow field is simulated based on the micro flow channel network, the minimum fluid resistance is taken as the target, and the optimal inclination of the pipe structure and the best flow direction of the lateral groove and other parameters are calculated. At the same time, the first channel derives intermediate features such as flow field uniformity coefficient and pressure distribution standard deviation in the analysis, which reflect the stability of the flow field under conventional adjustment.

[0100] After the second condition symbol is introduced into the second channel, the channel starts the directional response mechanism, combines the first trigger condition (such as the eddy current generator starts when the flow rate is less than 0.3 m / s) and the second trigger condition (such as the elastic connecting piece stretches when the pressure is greater than 50 Pa), and calculates the trigger parameters such as the rotational flow level of the micro eddy current generator and the stretching amount of the elastic clamping connecting piece. During the analysis process, the second channel generates trigger urgency index, component response delay and other intermediate features, reflecting the processing priority in special scenarios.

[0101] In the parallel decision-making process, the two channels realize lateral interaction of intermediate features through a high-speed data bus. The first channel transmits the flow field uniformity coefficient to the second channel to assist it in determining whether to reduce the eddy current intensity to avoid flow field disorder; the second channel feeds back the trigger urgency index to the first channel, and if the index is greater than 0.8, the first channel will automatically reduce the inclination adjustment range to reserve adjustment space for directional triggering.

[0102] Finally, the decision results of the two channels are integrated to generate a stent control strategy containing regular adjustment parameters and directional triggering instructions. For example, the strategy contains not only the regular adjustment of the inclination angle of the tubular structure 8° and the flow direction of the transverse slot 2°, but also the directional control of the 4-level start of the micro eddy current generator and the stretching of the elastic connecting piece by 18 mm, realizing the cooperation of two-dimensional control.

[0103] Through the classification of condition symbols, parallel decision-making of double channels and interaction of intermediate features, the generated stent control strategy can not only meet the stable adjustment needs in the regular environment, but also cope with the directional processing needs in special scenarios, improving the adaptability and accuracy of the strategy.

[0104] Further, the step A300 in the method provided in the embodiment of the application comprises:

[0105] A360: identifying the stent control strategy to perform strategy decomposition based on component driving of the M-type stent structure, and determining a sub-strategy set, wherein the sub-strategy set is identified with position coding.

[0106] A370: introducing a space-time code constraint, adding a constraint identifier to the sub-strategy set, distributing through the guide flow system, and performing component control management of the M-type stent structure, wherein the space-time code comprises a time stamp constraint and a position code constraint.

[0107] In one embodiment, when the M-type support structure is regulated, first, the support control strategy is identified. The support control strategy includes conventional adjustment parameters (such as the inclination angle of the tubular structural member 10°, the lateral groove flow direction 3°) and directional trigger instructions (such as the 3-level start of the micro vortex generator, the stretching of the elastic clamping connector by 15 mm), and the component control target, adjustment amplitude and associated conditions are extracted by the strategy analysis module to form a structured strategy data, and the analysis accuracy needs to reach more than 99% to ensure the accuracy of subsequent decomposition.

[0108] Then, based on the component driving of the M-type support structure, the overall strategy is split by component type. For the tubular structural member, the inclination adjustment sub-strategy is decomposed, such as the component adjustment of position code 001-050 to 10°, and the adjustment of 051-100 to 12°; for the lateral groove, the flow direction adjustment sub-strategy is decomposed, the flow direction of position 101-150 is 3°, and the flow direction of 151-200 is 5°; for the auxiliary component, the vortex level and stretching amount adjustment sub-strategy is decomposed, the vortex generator of position 020-030 is started at level 3, and the elastic connector of position 080-090 is stretched by 15 mm. Each sub-strategy is identified with the corresponding position code to ensure accurate association with the specific component, and the integrity of the sub-strategy set needs to cover all components that need to be regulated without omission.

[0109] Then, the time and space code constraints are introduced to add constraint identifiers to each sub-strategy. The timestamp in the time and space code is accurate to the second, such as 20250709153005, which represents the execution at 15:30:05 on July 9, 2025. According to the priority setting of component regulation, the time stamp deviation of emergency trigger type sub-strategy (such as vortex generator start) is controlled within ±0.1 seconds, and the deviation of conventional adjustment type (such as inclination adjustment) is not more than ±1 second; the position code constraint is consistent with the position code of the sub-strategy, such as the sub-strategy of position 001 corresponds to position code 001, ensuring that the constraint identifier matches the spatial properties of the sub-strategy.

[0110] Finally, the sub-strategy set with constraint identifiers is distributed by the diversion system. The diversion system uses a distributed communication protocol and transmits by position code partition, such as the sub-strategy of position 001-100 is transmitted through the first communication link, 101-200 through the second link, and the transmission delay is controlled within 0.5 seconds. After receiving the corresponding sub-strategy, the control module of each component executes at the specified time according to the timestamp constraint, for example, the vortex generator of position 020 starts 3-level rotational flow at 15:30:05, and the tubular structural member of position 001 is adjusted to 10° at 15:30:06, achieving precise control and management of components.

[0111] By identifying strategies, decomposing sub-strategies, adding spatiotemporal constraints, and distributing them in a distributed manner, precise and orderly control of each component of the M-type support structure was achieved, ensuring that the control strategy was effectively executed according to time and space requirements, and improving the timeliness and accuracy of support control.

[0112] In summary, the intelligent control method for rainwater drainage of an M-shaped bracket for photovoltaic panel roofs provided in this application has the following technical effects:

[0113] This application deploys an M-shaped support structure by acquiring the arrangement of photovoltaic panels, defining a tensor coding library, and deploying a first threshold at the data interface of the flow guidance system. Conditional symbols are generated based on feedback from multi-source environmental sensors. Then, the fluid computing engine is activated and the microchannel network is initialized based on the conditional symbols. The support control strategy is determined by conventional flow guidance and directional synaptic optimization, and the M-shaped support structure is regulated to achieve intelligent drainage of rainwater from the M-shaped support structure on the photovoltaic panel roof. This makes the rainwater drainage effect more precise and efficient, achieving the technical effect of accurately adapting to the rainwater drainage of the M-shaped support structure on the photovoltaic panel roof, realizing dynamic regulation, and improving the rainwater drainage effect.

[0114] Example 2, as Figure 2 As shown, based on the same inventive concept as in Embodiment 1 above, this application provides an intelligent control system for rainwater drainage of M-shaped brackets on photovoltaic panel roofs, the system comprising:

[0115] M-type support structure deployment module 1 is used to obtain the arrangement of photovoltaic panels and deploy the M-type support structure. The M-type support structure is a multi-layer architecture, with each layer including tubular structural components, horizontal grooves, vertical grooves and auxiliary components, and the layers are connected by elastic pipelines.

[0116] Conditional code acquisition module 2 is used to define a tensor coding library, deploy a first threshold on the data interface of the flow guidance system, and generate conditional codes by performing environmental tensor decomposition and flow guidance element transformation with the feedback of multi-source environmental sensors, wherein the conditional code is an element degree of freedom constraint.

[0117] M-type support structure control module 3 is used to activate the fluid computing engine deployed in the flow guidance system and initialize the built-in microchannel network according to the condition code, and determine the support control strategy by using the first conventional flow guidance and the second directional synapse optimization to regulate the M-type support structure, wherein the microchannel network is deployed based on the M-type support structure.

[0118] Furthermore, the condition code element acquisition module 2 is used to perform the following steps:

[0119] The tensor decomposition and weight filtering are performed along the environmental dimension to construct a first tensor matrix, wherein the weight is defined by the dredging correlation degree; the adjustable element is defined according to the component architecture of the M-type support structure to determine a second flow guide element matrix, wherein the adjustable element is identified by position coding; and the tensor code library is determined according to the first tensor matrix and the second flow guide element matrix.

[0120] Further, the conditional symbol acquisition module 2 is used to perform the following steps:

[0121] The encoding mode is introduced, wherein the encoding mode is a unified encoding mode for isomeric elements; the tensor code library is determined by performing matrix encoding through the encoding mode by concatenating the first tensor matrix and the second tensor matrix.

[0122] Further, the M-type support structure regulation module 3 is used to perform the following steps:

[0123] The micro-channel network is a twin flow field network of the M-type support structure, wherein the micro-channel network is consistent with the network state of the M-type support structure; and the fluid calculation engine is constructed according to the micro-channel network.

[0124] Further, the M-type support structure regulation module 3 is used to perform the following steps:

[0125] The first channel is supervised and trained through the built-in micro-channel network for the conventional scene flow guide decision and optimization based on the conditional symbol; the second channel is supervised and trained through the built-in micro-channel network for the directional conditional trigger decision and optimization based on the conditional symbol; the first channel and the second channel are concatenated, and lateral interaction is established to determine the fluid calculation engine.

[0126] Further, the M-type support structure regulation module 3 is used to perform the following steps:

[0127] The first trigger condition is set for the micro vortex generator, and the second trigger condition is set for the elastic clamping connector; and the second channel is constructed according to the first trigger condition and the second trigger condition.

[0128] Further, the M-type support structure regulation module 3 is used to perform the following steps:

[0129] The conditional symbol is identified to be partitioned at a preset distinction degree to determine the partition state of the M-type support structure, wherein each partition corresponds to a group of partition conditional symbols; if the number of partitions is greater than 1, the calculation domain is divided and the computing power is allocated according to the partition state to determine a calculation network, wherein the calculation network includes a plurality of parallel calculation nodes.

[0130] Further, the M-type support structure regulation module 3 is configured to perform the following steps:

[0131] The condition symbols are identified and partitioned with a preset distinction degree, and a partition state of the M-type support structure is determined, wherein each partition corresponds to a group of partition condition symbols; if the number of partitions is greater than 1, the calculation network is determined by performing calculation domain division and computing power allocation on the fluid calculation engine according to the partition state, wherein the calculation network comprises a plurality of parallel calculation nodes.

[0132] Further, the M-type support structure regulation module 3 is configured to perform the following steps:

[0133] The support control strategy is identified, the strategy decomposition is performed based on the component driving of the M-type support structure, and a sub-strategy set is determined, wherein the sub-strategy set is identified with position coding; the space-time code constraint is introduced, the constraint identifier is added to the sub-strategy set, the distributed deployment is performed through the flow guide system, and the component control management of the M-type support structure is performed, wherein the space-time code comprises a timestamp constraint and a position code constraint.

[0134] The photovoltaic panel shed M-type support rainwater drainage intelligent control system provided by the embodiment of the application can perform the photovoltaic panel shed M-type support rainwater drainage intelligent control method provided by any embodiment of the application, and has the corresponding function modules and beneficial effects of the execution method.

[0135] Although the present application makes various references to certain modules in the system according to the embodiments of the present application, however, any number of different modules can be used and run on the user terminal and / or server, and each unit and module included is only divided according to the functional logic, but is not limited to the above division, as long as the corresponding functions can be realized; in addition, the specific names of each functional unit are only for easy mutual differentiation, and do not limit the protection scope of the present application.

[0136] The above specific embodiments do not constitute a limitation on the protection scope of the present application. Those skilled in the art should understand that various modifications, combinations and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present application shall be included in the protection scope of the present application. In some cases, the actions or steps described in the present application can be performed in an order different from that in the embodiments and still achieve the desired results. In addition, the processes depicted in the drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multi-task processing and parallel processing are possible or can be advantageous.

Claims

1. A method for intelligent control of rainwater drainage on an M-shaped bracket for photovoltaic panel roofs, characterized in that, The method includes: Obtain the arrangement of photovoltaic panels and deploy an M-shaped support structure, wherein the M-shaped support structure is a multi-layer architecture, the layers of which include tubular structural components, transverse grooves, longitudinal grooves and auxiliary components, and the layers are connected by elastic pipelines. By defining a tensor coding library, a first threshold is deployed at the data interface of the flow guidance system. With the feedback from multi-source environmental sensors, environmental tensor decomposition and flow guidance element transformation are performed to generate conditional symbols, wherein the conditional symbols are element degree of freedom constraints. Based on the condition code, the fluid computing engine deployed in the flow guidance system is activated and the built-in microchannel network is initialized. The first conventional flow guidance and the second directional synapse optimization are used to determine the support control strategy and regulate the M-shaped support structure. The microchannel network is deployed based on the M-shaped support structure. The stent control strategy is determined by optimizing the first conventional diversion and the second directional synapse, including: Based on the conditional code, code binary classification is performed using the conventional flow-guiding dimension and the directional synapse dimension to determine the first conditional code and the second conditional code; The first condition code is imported into the first channel, the second condition code is imported into the second channel, parallel decision-making and lateral interaction of intermediate features are performed, and the support control strategy is output.

2. The method as described in claim 1, characterized in that, Define a tensor coding library, including: Tensor decomposition and weight filtering are performed on each environmental dimension to construct the first tensor matrix, where the weights are defined by the degree of correlation. For the component architecture of the M-type support structure, adjustable elements are defined, and a second flow guiding element matrix is ​​determined, wherein the adjustable elements are identified by position codes; The tensor encoding library is determined based on the first tensor matrix and the second flow element matrix.

3. The method as described in claim 2, characterized in that, The tensor encoding library is determined based on the first tensor matrix and the second flow element matrix, including: An encoding mode is introduced, wherein the encoding mode is a unified encoding method for homogeneous heterogeneous elements; By concatenating the first tensor matrix and the second tensor matrix, and using the aforementioned encoding mode, the tensor encoding library is determined through matrix encoding.

4. The method as described in claim 1, characterized in that, The microchannel network is a twin flow field network of an M-shaped support structure, wherein the microchannel network has the same network state as the M-shaped support structure. A fluid computing engine is constructed based on the microchannel network.

5. The method as described in claim 4, characterized in that, Based on the microchannel network, a fluid computing engine is constructed, including: By incorporating the microchannel network, the first channel is trained under supervision for routine scenario flow guidance decisions and optimization based on conditional symbols. The second channel is trained by using the built-in microchannel network to make decisions and optimize based on directional conditions triggered by conditional symbols. The first channel and the second channel are cascaded and lateral interaction is established to determine the fluid computing engine.

6. The method as described in claim 5, characterized in that, The auxiliary components include at least a micro eddy current generator and a flexible mounting connector; A first trigger condition is set for the micro eddy current generator, and a second trigger condition is set for the elastic mounting connector; The second channel is constructed based on the first triggering condition and the second triggering condition.

7. The method as described in claim 6, characterized in that, After activating the fluid computing engine, the method further includes: The condition code is identified, and partitioning is performed with a preset distinguishability to determine the partitioning status of the M-type support structure, wherein each partition corresponds to a set of partitioning condition code. If the number of partitions is greater than 1, the fluid computing engine is divided into computing domains and computing power is allocated according to the partition status to determine the computing network, wherein the computing network contains multiple parallel computing nodes.

8. The method as described in claim 1, characterized in that, Adjusting the M-shaped support structure includes: The support control strategy is identified, and the strategy is decomposed based on the component drive of the M-type support structure to determine the sub-strategy set, wherein the sub-strategy set is identified by a position code; Spatiotemporal code constraints are introduced, and constraint identifiers are added to the sub-strategy set. The strategy is distributed and deployed through the flow guidance system to perform component control and management of the M-type support structure. The spatiotemporal code includes timestamp constraints and location code constraints.

9. A smart control system for rainwater drainage of an M-shaped bracket for a photovoltaic panel roof, characterized in that, The system is used to implement the intelligent control method for rainwater drainage of an M-type bracket for photovoltaic panel roofs according to any one of claims 1-8, the system comprising: The M-type support structure deployment module is used to obtain the arrangement of photovoltaic panels and deploy the M-type support structure. The M-type support structure is a multi-layer architecture, with each layer including tubular structural components, horizontal grooves, vertical grooves and auxiliary components, and the layers are connected by elastic pipelines. The condition code acquisition module is used to deploy a first threshold on the data interface of the flow diversion system by defining a tensor coding library, and perform environmental tensor decomposition and flow diversion element transformation with the feedback of multi-source environmental sensors to generate condition codes, wherein the condition code is an element degree of freedom constraint. The M-type support structure control module is used to activate the fluid computing engine deployed in the flow guidance system and initialize the built-in microchannel network according to the condition code. It determines the support control strategy by optimizing the first conventional flow guidance and the second directional synapse, and controls the M-type support structure. The microchannel network is deployed based on the M-type support structure.

Images