Deep foundation pit intelligent precipitation method and system based on Beidou communication and water level switch linkage

Abstract

The invention discloses a deep foundation pit intelligent precipitation method and system based on Beidou communication and water level switch linkage, and relates to the technical field of deep foundation pit intelligent precipitation control, and the method comprises the steps: collecting and analyzing the deep foundation pit water level, precipitation equipment operation data and seasonal precipitation prediction, transmitting the data to a remote control center in real time, and receiving an instruction. A Beidou system and a mathematical model are used for predicting water level changes, and a rainfall strategy is automatically adjusted. The Beidou communication and water level switch technology is combined, and the remote communication capability and intelligent control of the deep foundation pit dewatering system are improved. Through big data and an AI algorithm, the system can predict a water level change trend, adjust a strategy in advance, improve precipitation efficiency and safety, and provide a new thought for deep foundation pit engineering management. Due to the integration of seasonal rainfall prediction data, the system can better adapt to rainfall in different seasons, safety accidents are avoided, and water resources are reasonably utilized.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of intelligent dewatering control of deep foundation pit, more particularly to a deep foundation pit intelligent dewatering method and system based on Beidou communication and water level switch linkage. BACKGROUND

[0002] In deep foundation pit engineering, dewatering is a key link to ensure construction safety. The traditional system mainly relies on water level monitoring and simple linkage of dewatering equipment to realize drainage. Ordinary water level sensors are usually used for monitoring, data is transmitted through wired networks or conventional wireless networks, and dewatering equipment is started and stopped by relying on manual threshold setting or simple logic control. There is a lack of integration of meteorological data, and the dewatering strategy is usually adjusted passively according to real-time water level, without forming a proactive prevention and control mechanism based on prediction. Traditional communication methods (such as 4G and Wi-Fi) are easily disturbed by signals in complex deep foundation pit environments (such as underground space and remote areas), resulting in data transmission delay or interruption, which cannot realize real-time monitoring. The remote control capability of the traditional system is limited, especially in scenarios with poor network coverage, making it difficult to realize remote debugging and parameter optimization of the dewatering equipment, and the operation and maintenance cost is high. Ordinary water level sensors have weak sensitivity to subtle water level changes, making it difficult to accurately capture the dynamic fluctuations of the water level in the foundation pit, which may cause early warning delay. Starting the dewatering equipment passively according to the current water level threshold lacks the ability to predict water level trends, and cannot respond to dewatering risks in advance, resulting in low dewatering efficiency and poor safety. Without integrating meteorological prediction data, it is difficult to adjust the strategy in advance according to the characteristics of dewatering in different seasons such as rainy season and dry season, which may cause water accumulation in the foundation pit due to sudden heavy rain, or waste of water resources due to excessive dewatering.

[0003] Therefore, how to propose a deep foundation pit intelligent dewatering method and system based on Beidou communication and water level switch linkage, by integrating seasonal precipitation prediction data, to better adapt to different seasonal precipitation, predict water level trends, adjust strategies in advance, and improve dewatering efficiency and safety is a problem that needs to be solved by those skilled in the art. SUMMARY

[0004] Therefore, the present application provides a deep foundation pit intelligent dewatering method and system based on Beidou communication and water level switch linkage, which can better adapt to different seasonal precipitation, predict water level trends, adjust strategies in advance, and improve dewatering efficiency and safety by integrating seasonal precipitation prediction data. In order to achieve the above purpose, the present application adopts the following technical solutions:

[0005] A deep foundation pit intelligent dewatering method based on Beidou communication and water level switch linkage, comprising:

[0006] Collecting water level data, dewatering equipment operation data and seasonal precipitation prediction data of the deep foundation pit;

[0007] According to the water level data of the deep foundation pit and the pre-set rules, the dewatering equipment is controlled to dewater;

[0008] The collected data is transmitted to the remote control center in real time by using the Beidou satellite navigation system, and instructions from the control center are received.

[0009] The collected water level data of the deep foundation pit, the operation data of the dewatering equipment and the seasonal precipitation prediction data are comprehensively analyzed, a mathematical model is established, the change trend of the water level in the deep foundation pit is predicted, and the dewatering strategy is adjusted.

[0010] According to the dewatering strategy, the opening and closing of the dewatering equipment and the adjustment of the dewatering intensity are automatically controlled.

[0011] Optionally, the water level data of the deep foundation pit is collected by using a water level switch as a core sensor to accurately perceive the subtle changes of the water level in the foundation pit, and the water level switch sets multiple different water level thresholds, such as an alarm water level and a dangerous water level.

[0012] Optionally, the collected data is transmitted to the remote control center by using the Beidou short message communication mode, and the water level data and the operation data of the dewatering equipment are transmitted to the receiving equipment of the remote control center or the relevant staff, and instructions from the control center are received.

[0013] Optionally, according to the water level data of the deep foundation pit and the pre-set rules, the opening and closing of the dewatering equipment and the adjustment of the dewatering intensity are automatically controlled, when the water level reaches the alarm water level, part of the water pumps are automatically started to dewater, and when the water level continuously rises to the dangerous water level, the working power of the water pumps is increased or more water pumps are started to speed up the dewatering speed.

[0014] Optionally, before the comprehensive analysis of the collected water level data of the deep foundation pit, the operation data of the dewatering equipment and the seasonal precipitation prediction data, the following steps are included:

[0015] The seasonal precipitation prediction data provided by the meteorological department, the water level data collected by the system and the geological data around the foundation pit are normalized, and the data is mapped to the [0, 1] interval, and the formula is as follows:

[0016]

[0017] Wherein, X is the original data, X min and X max are the minimum and maximum values of the data characteristics, and X norm is the normalized data.

[0018] Optionally, the establishing a mathematical model and predicting the change trend of the water level in the deep foundation pit comprises:

[0019] Based on LSTM, multi-source data feature extraction and fusion are performed. When processing time series data, the long short-term memory network LSTM captures long-term dependencies.

[0020] The normalized multi-source data is divided by time step T to construct an input matrix.

[0021] Suppose the input data is [P t , Q t , F, W t , K, φ], which is input into the LSTM network, wherein P t is a precipitation time series, Q t is a precipitation amount sequence, F t is a precipitation frequency sequence, W t is water level data collected by the system, K is a permeability coefficient, and φ is a porosity.

[0022] Through layer-by-layer processing of the multi-layer LSTM network, a fused feature vector H is finally obtained, which contains the internal relationship features between the multi-source data.

[0023] Optionally, it further comprises: based on random forest, data feature importance analysis and fusion optimization are performed, the fused feature vector H and the corresponding foundation pit water level change label are taken as the input of the random forest, the importance score of each feature is obtained by calculating the contribution of each feature to reducing impurity when the decision tree node is split, and the data is weighted and fused according to the importance score, and the formula is as follows:

[0024]

[0025] Wherein, F final is the final fusion result, w i is the weight of the i-th feature, h i is the i-th feature value, and n is the number of features.

[0026] Optionally, the adjusting the dewatering strategy comprises: performing dewatering equipment parameter dynamic adjustment based on Kalman filtering, dynamically adjusting the operation parameters of the dewatering equipment according to the seasonal precipitation prediction information, and setting the state equation of the foundation pit water level as:

[0027] x t =Ax t-1 +Bu t +w t ;

[0028] The observation equation is: z t =Hx t +vt ;

[0029] wherein, x t is the state vector of the foundation pit water level at time t, A is the state transition matrix, B is the control matrix, u t is the control input, w t is the process noise, z t is the observation vector, H is the observation matrix, v t is the observation noise;

[0030] The prediction step of Kalman filtering:

[0031]

[0032] P t|t-1 = AP t-1|t-1 A T +Q;

[0033] The update step:

[0034] K t =P t|t-1 H T (HP t|t-1 H T +R) -1 ;

[0035]

[0036] P t|t =(I-K t H)P t|t-1 ;

[0037] wherein, is the prediction of the state at time t based on the state at time t-1, is the state estimation at time t, P t|t-1 is the prediction error covariance, P t|t is the updated error covariance, K t is the Kalman gain, by continuously iterating the above process, according to the real-time rainfall and rainfall prediction update, dynamically adjusting the operation parameters u t of the dewatering equipment, controlling the foundation pit water level.

[0038] Optionally, it also includes performing reinforcement learning-based dewatering strategy optimization, modeling the foundation pit dewatering process as a Markov decision process, defining the state space S, the action space A, and the reward function R, using the deep Q network algorithm for policy learning, approximating the Q function through the neural network, and updating the neural network parameters according to the reward feedback during the training process to learn the optimal dewatering strategy for effective control of the foundation pit water level under different dewatering conditions.

[0039] Optionally, a deep foundation pit intelligent dewatering system based on Beidou communication and water level switching linkage comprises:

[0040] The acquisition module is used for collecting water level data, dewatering equipment operation data and seasonal dewatering prediction data of the deep foundation pit.

[0041] The dewatering control module is used for controlling the dewatering equipment to dewater according to the water level data of the deep foundation pit and a preset rule.

[0042] The Beidou communication module is used for transmitting the collected data to a remote control center in real time by using the Beidou satellite navigation system and receiving instructions from the control center.

[0043] The strategy adjustment module is used for comprehensively analyzing the collected water level data, dewatering equipment operation data and seasonal dewatering prediction data of the deep foundation pit, establishing a mathematical model, predicting the change trend of the water level in the deep foundation pit and adjusting the dewatering strategy.

[0044] The dewatering control adjustment module is used for automatically controlling the opening, closing and dewatering intensity adjustment of the dewatering equipment according to the dewatering strategy.

[0045] According to the above technical solution, compared with the prior art, the present application provides a deep foundation pit intelligent dewatering method and system based on Beidou communication and water level switching linkage, which has the following beneficial effects:

[0046] The application provides a deep foundation pit intelligent precipitation method based on Beidou communication and water level switch linkage, which comprises the following steps: collecting water level data of a deep foundation pit, precipitation equipment operation data and seasonal precipitation prediction data; controlling the precipitation equipment to carry out precipitation according to the water level data of the deep foundation pit and a pre-set rule; transmitting the collected data to a remote control center in real time by using a Beidou satellite navigation system, and receiving instructions from the control center; comprehensively analyzing the collected water level data of the deep foundation pit, the precipitation equipment operation data and the seasonal precipitation prediction data, establishing a mathematical model, predicting the water level change trend in the deep foundation pit, and adjusting the precipitation strategy; and automatically controlling the opening, closing and precipitation intensity adjustment of the precipitation equipment according to the precipitation strategy. The Beidou communication technology is combined with the water level switch and applied to the deep foundation pit intelligent precipitation system. The Beidou communication technology solves the problem of limited data transmission in the traditional deep foundation pit precipitation system, so that the system can realize remote and stable data communication in a complex environment. The high-precision monitoring capability of the water level switch and the real-time transmission capability of the Beidou communication are matched with each other, thereby providing a solid technical foundation for the intelligent control of deep foundation pit precipitation. The intelligent decision module based on big data analysis and artificial intelligence algorithm can predict the water level change trend in the deep foundation pit and adjust the precipitation strategy in advance according to the seasonal precipitation prediction, by comprehensively analyzing and processing multi-source data. Compared with the traditional passive precipitation control method, the active and intelligent decision mode can greatly improve the efficiency and safety of deep foundation pit precipitation, and brings a new idea and method for the precipitation management of deep foundation pit engineering. The seasonal precipitation prediction data is integrated into the control process of the deep foundation pit intelligent precipitation system, the precipitation strategy is adjusted in advance according to the characteristics of seasonal precipitation, the system can better cope with the influence of precipitation in different seasons on the deep foundation pit, effectively avoid the safety accidents of the deep foundation pit caused by precipitation, and realize the reasonable utilization of water resources. BRIEF DESCRIPTION OF DRAWINGS

[0047] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings needed to be used in the embodiments or prior art description. Obviously, the drawings in the following description only constitute the embodiments of the present application, and for those skilled in the art, other drawings can also be obtained without creative labor on the basis of the provided drawings.

[0048] Figure 1 A deep foundation pit intelligent precipitation method based on Beidou communication and water level switch linkage provided by the present application is shown in the flowchart.

[0049] Figure 2 A deep foundation pit intelligent precipitation system structure framework based on Beidou communication and water level switch linkage provided by the present application is shown in the block diagram. DETAILED DESCRIPTION

[0050] Clearly, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative efforts fall within the protection scope of the present application.

[0051] The embodiment of the present application discloses a deep foundation pit intelligent dewatering method based on Beidou communication and water level switch linkage, as shown in the figure, comprising: Figure 1

[0052] Collecting water level data, dewatering equipment operation data and seasonal dewatering prediction data of the deep foundation pit;

[0053] Controlling the dewatering equipment to dewater according to the water level data of the deep foundation pit and the pre-set rules;

[0054] Using the Beidou satellite navigation system to perform real-time transmission of data, sending the collected data to a remote control center, and receiving instructions from the control center;

[0055] Comprehensively analyzing the collected water level data, dewatering equipment operation data and seasonal dewatering prediction data of the deep foundation pit, establishing a mathematical model, predicting the change trend of the water level in the deep foundation pit, and adjusting the dewatering strategy;

[0056] According to the dewatering strategy, automatically controlling the opening, closing and dewatering intensity adjustment of the dewatering equipment.

[0057] Further, the collection of water level data of the deep foundation pit includes: using a water level switch as a core sensor to accurately perceive the subtle changes of the water level in the foundation pit, and the water level switch sets multiple different water level thresholds, such as warning water level and dangerous water level.

[0058] Further, the sending of the collected data to the remote control center includes: sending the water level data, dewatering equipment operation data of the deep foundation pit to the receiving equipment of the remote control center or relevant staff through the Beidou short message communication mode, and receiving instructions from the control center.

[0059] Further, the control of the dewatering equipment to dewater according to the water level data of the deep foundation pit and the pre-set rules includes: automatically controlling the opening, closing and dewatering intensity adjustment of the dewatering equipment according to the water level data of the deep foundation pit and the pre-set rules, automatically starting part of the water pump to dewater when the water level reaches the warning water level; when the water level continuously rises to the dangerous water level, increasing the working power of the water pump or starting more water pumps to speed up the dewatering speed.

[0060] ​Further, the collected water level data of the deep foundation pit, the dewatering equipment operation data and the seasonal precipitation prediction data are comprehensively analyzed before including:

[0061] The seasonal precipitation prediction data provided by the meteorological department: P t is a precipitation time series, Q t is a precipitation amount sequence, F t is a precipitation frequency sequence, W t is water level data collected by the system, K is a permeability coefficient, and φ is a porosity.

[0062]

[0063] Among them, X is the original data, X min and X max are the minimum and maximum values of the data characteristics respectively, and X norm is the normalized data.

[0064] Further, the mathematical model is established to predict the change trend of the water level in the deep foundation pit, including:

[0065] Based on LSTM, multi-source data feature extraction and fusion are performed. When processing time series data, the long short-term memory network LSTM captures long-term dependencies.

[0066] The normalized multi-source data is divided according to the time step T to construct an input matrix.

[0067] Let the input data be [P t , Q t , F, W t , K, φ], which is input into the LSTM network, wherein P t is a precipitation time series, Q t is a precipitation amount sequence, F t is a precipitation frequency sequence, W t is water level data collected by the system, K is a permeability coefficient, and φ is a porosity.

[0068] Specifically, the calculation formulas of the forget gate, the input gate and the output gate in the LSTM network are as follows:

[0069] Forget gate:

[0070] f t =σ(W f ·[h t-1 ,x t ]+b f );

[0071] Input gate:

[0072] i t = σ(W i · [h t-1 , x t ] + b i ) ;

[0073]

[0074] Cell state update:

[0075]

[0076] Output gate:

[0077] o t = σ(W o · [h t-1 , x t ] + b o ) ;

[0078] h t = o t * tanh(C t ) ;

[0079] where σ is a sigmoid activation function, W f , W i , W C , W o are weight matrices, b f , b i , b C , b o are bias vectors, h t-1 is the hidden state at the previous time, x t is the input data at the current time, f t , i t , o t are the outputs of the forget gate, input gate, and output gate, respectively, is the candidate cell state, C t is the updated cell state, and h t is the hidden state at the current time.

[0080] Through layer-by-layer processing of the multi-layer LSTM network, a fused feature vector H is finally obtained, which contains the intrinsic relationship features between the multi-source data.

[0081] Through layer-by-layer processing of the multi-layer LSTM network, a fused feature vector H is finally obtained, which contains the intrinsic relationship features between the multi-source data.

[0082] Further, it also includes: based on random forest, data feature importance analysis and fusion optimization, taking the fused feature vector H and the corresponding foundation pit water level change label as the input of the random forest, obtaining the importance score of each feature by calculating the contribution of each feature to reducing impurity when the decision tree node is split, and according to the importance score, the data is weighted and fused, and the formula is as follows:

[0083]

[0084] Where F final is the final fusion result, w i is the weight of the i-th feature, determined according to the importance score, h i is the i-th feature value, and n is the number of features.

[0085] Further, the adjustment of the precipitation strategy includes: performing dynamic adjustment of precipitation equipment parameters based on Kalman filtering, dynamically adjusting the operation parameters of the precipitation equipment according to the seasonal precipitation prediction information, and setting the state equation of the foundation pit water level as:

[0086] x t =Ax t-1 +Bu t +w t ;

[0087] The observation equation is: z t =Hx t +v t ;

[0088] Where x t is the state vector of the foundation pit water level at time t (including water level height, water level change rate and other information), A is the state transition matrix, B is the control matrix, u t is the control input (such as the opening time of the precipitation equipment, the pumping rate, etc.), w t is the process noise (complying with the Gaussian distribution with mean 0 and covariance Q), z t is the observation vector (actual measured water level data), H is the observation matrix, v t is the observation noise (complying with the Gaussian distribution with mean 0 and covariance R);

[0089] The prediction step of Kalman filtering is:

[0090]

[0091] P t|t-1 =AP t-1|t-1 A T +Q;

[0092] The update step is:

[0093] K t =P t|t-1 H T (HP t|t-1 H T +R) -1 ;

[0094]

[0095] P t|t =(I-K t H)P t|t-1 ;

[0096] where, is the prediction of state at time t based on state at time t-1, is the state estimation at time t, P t|t-1 is the prediction error covariance, P t|t is the updated error covariance, K t is the Kalman gain, by continuously iterating the above process, according to the real-time rainfall and rainfall prediction update, dynamically adjust the operation parameters u of the dewatering equipment t , control the water level of the foundation pit.

[0097] Further, it also includes performing reinforcement learning-based dewatering strategy optimization, modeling the foundation pit dewatering process as a Markov decision process (MDP), defining the state space S (including foundation pit water level, dewatering data, geological data, etc.), action space A (various combinations of operation parameters of dewatering equipment), reward function R (such as giving positive reward if water level is kept within safe range, and negative reward if it exceeds the range).

[0098] Specifically, the obtained state estimation is updated to the state space S' of reinforcement learning.

[0099] Under the new state space S', reinforcement learning selects action a∈A (various combinations of operation parameters of dewatering equipment) according to the current state. The design of reward function R takes into account whether the water level is kept within the safe range, and increases the consideration of the accuracy of Kalman filter estimation.

[0100] If the Kalman filter estimation error (such as ), within a certain range, give additional positive reward; if the error is too large, give negative reward.

[0101] Deep Q-network (DQN) algorithm is used for strategy learning, and neural network is used to approximate Q function:

[0102] Q(s′,a;θ)≈Q * (s′,a);

[0103] Wherein s' is the fused state, a is the action, and θ is the parameter of the neural network.

[0104] During the training process, the neural network parameters are updated according to the reward feedback by continuously interacting with the environment, and the optimal precipitation strategy is learned to effectively control the water level of the foundation pit under different precipitation conditions.

[0105] In the specific embodiment, in the learning process, in addition to optimizing its own strategy, the reinforcement learning also adjusts the parameters of the Kalman filter (process noise covariance Q, observation noise covariance R, etc.) according to the reward feedback obtained.

[0106] When the reward is high and the Kalman filter estimation is accurate, the current parameters are maintained; when the reward is low and the estimation error is found to be large, the values of Q and R are adjusted by the gradient descent method to improve the estimation accuracy of the Kalman filter.

[0107] The specific algorithm flow includes:

[0108] S1: Initialization: Initialize the parameters of the Kalman filter (A, B, H, Q, R), the neural network parameters θ of the reinforcement learning, the state space S', the action space A, and the reward function R.

[0109] S2: Data acquisition: Obtain the precipitation prediction data from the meteorological department, the water level data collected by the system, and the geological data around the foundation pit, etc.

[0110] S3: Kalman filter state update: According to the collected data, the state estimation and update are performed using the Kalman filter algorithm to obtain and update the state space S' of the reinforcement learning.

[0111] S4: Reinforcement learning decision: The reinforcement learning selects the action a through the DQN network according to the current state space S' to control the operation of the precipitation equipment.

[0112] S5: Environment interaction and reward calculation: After executing the action a, interact with the environment, obtain new observation data, and calculate the reward R, including the water level state reward and the Kalman filter estimation accuracy reward.

[0113] S6: Parameter update: Update the neural network parameters θ of the reinforcement learning using the reward R and the new state; at the same time, according to the reward situation, optimize and adjust the parameters of the Kalman filter.

[0114] S7: Repeat iteration: Repeat steps S2-S6 until the preset training number is reached or the convergence condition is met, and obtain the optimal precipitation strategy and Kalman filter parameters. The above fusion scheme of the embodiment complements the advantages of the two algorithms and improves the control effect of the foundation pit precipitation.

[0115] In the specific embodiment, a deep foundation pit intelligent dewatering system based on Beidou communication and water level switch linkage comprises the following parts as shown in the figure: Figure 2

[0116] The collection module is used for collecting water level data of the deep foundation pit, dewatering equipment operation data and seasonal dewatering prediction data.

[0117] The dewatering control module is used for controlling the dewatering equipment to dewater according to the water level data of the deep foundation pit and pre-set rules.

[0118] The Beidou communication module is used for real-time transmission of data by using the Beidou satellite navigation system, sending the collected data to the remote control center and receiving instructions from the control center.

[0119] The strategy adjustment module is used for comprehensive analysis of the collected water level data of the deep foundation pit, dewatering equipment operation data and seasonal dewatering prediction data, establishment of a mathematical model, prediction of the water level change trend in the deep foundation pit and adjustment of the dewatering strategy.

[0120] The dewatering control adjustment module is used for automatic control of the opening, closing and dewatering intensity adjustment of the dewatering equipment according to the dewatering strategy.

[0121] The water level and dewatering are linked in real time, once the water level switch detects that the water level change reaches the set threshold, the signal is triggered through the internal circuit and transmitted to the dewatering control module. After receiving the signal, the dewatering control module quickly performs corresponding control operation on the dewatering equipment, realizes the real-time linkage of the water level and dewatering, and ensures that the water level in the foundation pit is always within the safe range. In the whole linkage process, the Beidou communication module is responsible for real-time transmission of water level monitoring data, dewatering equipment state data and linkage operation record information to the remote control center. At the same time, parameter adjustment instructions and remote control instructions sent by the control center are received, and the remote monitoring and management of the system are ensured. Even in the case that the network signal at the location of the foundation pit is not good, the Beidou communication can stably complete the data transmission task and ensure the normal operation of the system. The seasonal dewatering prediction data provided by the meteorological department, such as dewatering time, dewatering amount and dewatering frequency, are fused with the water level data, surrounding geological data and other data collected by the system. Through data mining and analysis technology, the internal relationship between seasonal dewatering and water level change of the foundation pit is found out, and more comprehensive data support is provided for intelligent decision-making. Before the rainy season, the intelligent decision-making module adjusts the dewatering strategy in advance according to the seasonal dewatering prediction information. For example, the opening time of the dewatering equipment is increased, the initial water level in the foundation pit is reduced, and enough space is reserved to cope with possible heavy dewatering. During the dewatering process, the operation parameters of the dewatering equipment are dynamically adjusted according to the real-time dewatering condition and the update of the dewatering prediction, so as to realize accurate control of the water level in the foundation pit.

[0122] ​The various embodiments described in this specification are implemented in a progressive manner, each embodiment focusing on the differences from other embodiments, and the same or similar parts between embodiments can be mutually referred to. For the apparatus disclosed by the embodiments, since it corresponds to the method disclosed by the embodiments, the description is relatively simple, and the relevant parts can be referred to the description of the method part.

[0123] The above description of disclosed embodiments enables one of ordinary skill in the art to make or use the application. Various modifications to these embodiments will be readily apparent to those of ordinary skill in the art, and the generic principles defined herein can be applied to other embodiments without departing from the spirit or scope of the application. Therefore, the application is not intended to be limited to the embodiments shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A deep foundation pit intelligent dewatering method based on Beidou communication and water level switch linkage, characterized in that: include: Collect water level data of deep foundation pits, precipitation equipment operation data, and seasonal precipitation forecast data; Control the dewatering equipment to dewater according to the water level data of the deep foundation pit and pre-set rules; Utilize the BeiDou satellite navigation system for real-time data transmission, send the collected data to the remote control center, and receive instructions from the control center at the same time; Comprehensively analyze the collected deep foundation pit water level data, precipitation equipment operation data, and seasonal precipitation forecast data to establish a mathematical model to predict the changing trend of the deep foundation pit water level and adjust the precipitation strategy; Automatically control the opening and closing of precipitation equipment and the adjustment of precipitation intensity according to precipitation strategies.

2. The method for intelligent dewatering of deep foundation pits based on Beidou communication and water level switch linkage according to claim 1 is characterized in that: The method of collecting water level data of a deep foundation pit includes: using a water level switch as a core sensor to accurately sense subtle changes in the water level in the foundation pit, and setting multiple different water level thresholds on the water level switch, such as a warning water level and a dangerous water level.

3. The deep foundation pit intelligent dewatering method based on Beidou communication and water level switch linkage according to claim 1 is characterized in that: The sending of the collected data to the remote control center includes: sending the water level data of the deep foundation pit and the operation data of the precipitation equipment to the receiving device of the remote control center or relevant staff through the Beidou short message communication method, and at the same time, receiving instructions from the control center.

4. The method for intelligent dewatering of deep foundation pits based on Beidou communication and water level switch linkage according to claim 1 is characterized in that: The method of controlling the precipitation equipment to carry out precipitation according to the water level data of the deep foundation pit and the pre-set rules includes: automatically controlling the opening and closing of the precipitation equipment and the adjustment of the precipitation intensity according to the water level data of the deep foundation pit and the pre-set rules; when the water level reaches the warning level, automatically starting some water pumps to carry out precipitation; when the water level continues to rise to the dangerous level, increasing the working power of the water pump or starting more water pumps to speed up the precipitation speed.

5. The deep foundation pit intelligent dewatering method based on Beidou communication and water level switch linkage according to claim 1 is characterized in that: The comprehensive analysis of the collected deep foundation pit water level data, precipitation equipment operation data and seasonal precipitation forecast data includes: The seasonal precipitation forecast data provided by the meteorological department, the water level data collected by the system, and the geological data around the foundation pit are normalized and mapped to the [0, 1] interval. The formula is as follows: Among them, X is the original data, X min and X max are the minimum and maximum values ​​of the data feature, respectively, X norm The data are normalized.

6. The method for intelligent dewatering of deep foundation pits based on Beidou communication and water level switch linkage according to claim 1 is characterized in that: The mathematical model established to predict the changing trend of the water level in the deep foundation pit includes: Extract and fuse multi-source data features based on LSTM, and capture long-term dependencies when processing time series data through the long short-term memory network LSTM; The normalized multi-source data is divided according to the time step T to construct the input matrix; Assume the input data is [P t , Q t , F, W t ,K,φ], which is input into the LSTM network, where P t is the precipitation time series, Q t is the precipitation series, F t is the precipitation frequency series, W t is the water level data collected by the system, K is the permeability coefficient, and φ is the porosity; Through layer-by-layer processing of the multi-layer LSTM network, the fused feature vector H is finally obtained, which contains the intrinsic connection characteristics between multi-source data.

7. The method for intelligent dewatering of deep foundation pits based on Beidou communication and water level switch linkage according to claim 6 is characterized in that: Also includes: Based on random forest, data feature importance analysis and fusion optimization are performed. The fused feature vector H and the corresponding foundation pit water level change label are used as the input of random forest. By calculating the contribution of each feature to reducing impurity when the decision tree node is split, the importance score of each feature is obtained. According to the importance score, the data is weighted and fused. The formula is as follows: Among them, F final is the final fusion result, w i is the weight of the i-th feature, h i is the i-th eigenvalue, and n is the number of features.

8. The method for intelligent dewatering of deep foundation pits based on Beidou communication and water level switch linkage according to claim 1 is characterized in that: The precipitation adjustment strategy includes: Dynamic adjustment of precipitation equipment parameters based on Kalman filtering is performed. According to seasonal precipitation forecast information, the Kalman filtering algorithm is used to dynamically adjust the operating parameters of precipitation equipment. The state equation of the foundation pit water level is set as: x t =Ax t-1 +Bu t +w t ; The observation equation is: t =Hx t +v t ; Among them, x t is the state vector of the foundation pit water level at time t, A is the state transfer matrix, B is the control matrix, u t is the control input, w t is the process noise, z t is the observation vector, H is the observation matrix, v t is the observation noise; Kalman filter prediction steps: P t|t-1 =AP t-1|t-1 From T +Q; Update steps: K t =P t|t-1 H T (HP t|t-1 H T +R) -1 ; P t|t =(I-K t H)P t|t-1 ; in, is the prediction of the state at time t based on the state at time t-1, is the state estimate at time t, P t|t-1 is the prediction error covariance, P t|t is the updated error covariance, K t is the Kalman gain. By continuously iterating the above process, the operating parameters u of the precipitation equipment are dynamically adjusted according to the real-time precipitation situation and the update of precipitation forecast. t , to control the water level of the foundation pit.

9. The method for intelligent dewatering of deep foundation pits based on Beidou communication and water level switch linkage according to claim 8 is characterized in that: It also includes the optimization of precipitation strategies based on reinforcement learning, modeling the foundation pit precipitation process as a Markov decision process, defining the state space S, action space A, and reward function R, using the deep Q network algorithm for strategy learning, and approximating the Q function through the neural network. During the training process, by constantly interacting with the environment, the neural network parameters are updated according to the reward feedback, and the optimal precipitation strategy is learned to effectively control the foundation pit water level under different precipitation conditions.

10. A deep foundation pit intelligent dewatering system based on Beidou communication and water level switch linkage, characterized in that: include: Collection module: used to collect water level data of deep foundation pits, precipitation equipment operation data and seasonal precipitation forecast data; Dewatering control module: used to control dewatering equipment to carry out dewatering according to the water level data of the deep foundation pit and pre-set rules; Beidou communication module: used to transmit data in real time using the Beidou satellite navigation system, send collected data to the remote control center, and receive instructions from the control center; Strategy Adjustment Module: This module is used to comprehensively analyze the collected deep foundation pit water level data, precipitation equipment operation data, and seasonal precipitation forecast data, establish a mathematical model, predict the changing trend of the deep foundation pit water level, and adjust the precipitation strategy; Precipitation control and adjustment module: used to automatically control the opening and closing of precipitation equipment and the adjustment of precipitation intensity according to precipitation strategies.

Images