Intelligent pipe control system and method for fluid pipe network based on multi-order space-time feature fusion

Abstract

The application provides a kind of fluid pipe network intelligent management and control system and method based on multi-order space-time feature fusion, comprising: embedded internet of things sensing layer, including embedded remote information transmission module electric valve control mainboard and / or pneumatic valve control mainboard, and with the flow sensor, heat sensor and water invasion sensor connected with the control mainboard;Multi-order space-time feature processing layer, comprising: space-time feature extraction module is configured to carry out sliding window slicing processing to the time series data of pressure, flow rate, temperature parameter;Feature fusion module is configured to carry out feature alignment to vibration spectrum, voiceprint signal and infrared thermal imaging data by multi-head attention mechanism.A kind of fluid pipe network intelligent management and control system and method based on multi-order space-time feature fusion described in the application, through embedded multi-source sensing, multi-order space-time feature fusion and intelligent decision architecture, significantly improve the management and control efficiency and safety of fluid pipe network.

Description

Technical Field

[0001] This invention belongs to the field of data processing technology, and in particular relates to an intelligent control system and method for fluid pipeline networks based on multi-level spatiotemporal feature fusion. Background Technology

[0002] Traditional fluid pipeline network management and control systems often rely on single sensor data and static models, making it difficult to effectively integrate multi-source heterogeneous spatiotemporal characteristics, resulting in delayed fault warnings and insufficient robustness of control strategies.

[0003] In existing technologies, the data acquisition frequency is low (usually below 1kHz), the feature extraction dimension is single (such as using only time domain statistics), and there is a lack of cross-platform real-time interaction capabilities. Especially when dealing with emergency situations such as sudden changes in pressure, the reliance on human decision-making is high, which can easily lead to misoperation.

[0004] In addition, traditional database architectures are difficult to support high-frequency concurrent data processing, pose significant cybersecurity risks, and cannot meet the real-time control requirements of industrial IoT scenarios. Summary of the Invention

[0005] In view of this, the present invention aims to propose an intelligent control system and method for fluid pipeline networks based on multi-level spatiotemporal feature fusion, so as to solve at least one of the problems in the background art.

[0006] To achieve the above objectives, the technical solution of the present invention is implemented as follows:

[0007] A fluid pipeline network intelligent control system based on multi-order spatiotemporal feature fusion includes:

[0008] An embedded IoT sensing layer includes an electric valve control motherboard and / or a pneumatic valve control motherboard with an embedded remote information transmission module, as well as a flow sensor, a heat sensor and a water intrusion sensor connected to the control motherboard.

[0009] Multi-level spatiotemporal feature processing layer, including:

[0010] The spatiotemporal feature extraction module is configured to perform sliding window slicing on time-series data of pressure, flow velocity, and temperature parameters to extract first-order statistical features including mean and variance, second-order frequency domain features of Fourier transform coefficients, and third-order nonlinear features of Lyapunov exponents.

[0011] The feature fusion module is configured to perform feature alignment on vibration spectrum, acoustic signature signal and infrared thermal imaging data through a multi-head attention mechanism;

[0012] The intelligent decision-making layer, wherein the output of the multi-stage spatiotemporal feature processing layer is connected to the input of the intelligent decision-making layer through a message queue middleware, includes:

[0013] The digital twin module constructs a visual model that includes the three-dimensional topology of the pipeline and fluid dynamics parameters;

[0014] The fault prediction module includes an LSTM neural network with residual connections.

[0015] The linkage control module is configured to generate an instruction queue containing valve identifiers and execution timing.

[0016] The cross-platform interaction layer includes a Linux server running a real-time in-memory database and a Windows terminal that communicates with the server via the OPC UA protocol.

[0017] Furthermore, in the embedded IoT sensing layer:

[0018] The remote information transmission module includes a 4G / 5G dual-mode communication chip and an AES-256 encryption unit;

[0019] The vibration sensor array acquires vibration waveforms at a sampling rate of 50kHz and is equipped with a wavelet packet transform processor;

[0020] The water intrusion sensor includes an impedance spectrum analysis circuit with an operating frequency range of 10-100kHz. The water intrusion sensor is connected to the control motherboard via an RS-485 bus to transmit the impedance spectrum phase angle change.

[0021] Furthermore, the digital twin modeling module:

[0022] An unsteady fluid simulation model was constructed using the finite volume method, and the solver was configured to execute the SIMPLE algorithm.

[0023] The dynamic boundary condition generation unit receives monitoring data from the pressure sensor and valve opening sensor in real time;

[0024] Includes a database of corrected seal friction coefficients based on the number of valve opening and closing cycles.

[0025] Furthermore, the fault prediction module:

[0026] The training dataset contains samples labeled with normal operating conditions and 5%-15% fault injection conditions;

[0027] The anomaly detection unit is configured to calculate the Mahalanobis distance between the real-time feature vector and the cluster center;

[0028] The output layer includes an adaptive threshold adjustment unit based on ambient temperature;

[0029] The residual connection is configured to span at least three time steps, and the skip links of the residual connection receive frequency domain feature input from the spatiotemporal feature extraction module.

[0030] Furthermore, the linkage control module:

[0031] Includes a valve priority sorting unit based on pipeline topology;

[0032] Instruction queues are divided into time-sensitive queues and preventative maintenance queues;

[0033] It is equipped with a command retransmission mechanism, and the retransmission trigger time threshold is 2 seconds. The 2-second threshold is determined based on a safety factor of 1.5 times the propagation speed of the pipeline pressure wave.

[0034] Furthermore, the cross-platform interaction layer:

[0035] The real-time in-memory database uses a Redis cluster architecture and includes millisecond-level data caching, feature intermediate result caching, and diagnostic result caching.

[0036] The network isolation unit includes a hardware firewall and a MAC address-based whitelist access controller;

[0037] The visualization terminal is equipped with a WebGL rendering engine and parallel pressure cloud viewport and feature heatmap viewport.

[0038] Furthermore, the network isolation unit:

[0039] Divide the IoT sensing layer into independent security domains and configure unidirectional communication channels;

[0040] Includes a Modbus TCP protocol parser, configured to filter write register instructions;

[0041] The control command transmission channel is equipped with a quantum key distribution unit, which provides dynamic key updates for the AES-256 encryption unit.

[0042] Furthermore, this solution discloses an intelligent management and control method for fluid pipeline networks based on multi-order spatiotemporal feature fusion, including:

[0043] Step S1: Collect valve vibration spectrum, fluid pressure waveform, and pipe wall temperature gradient data through embedded sensing nodes, with a sampling frequency of not less than 10kHz. The valve vibration spectrum is collected through the vibration sensor array.

[0044] Step S2: Using an improved empirical mode decomposition algorithm, the original signal is decomposed into intrinsic mode functions (IMFs), and the 2nd to 5th order IMF components are retained to reconstruct the characteristic signal.

[0045] Step S3: Construct a spatiotemporal feature cube, divide 60 consecutive seconds of data into 20 overlapping time windows, and extract 128-dimensional time domain features and 64-dimensional frequency domain features from each window;

[0046] Step S4: Train a multi-task learning model. The first task is to predict the location of the pipeline leak, the second task is to estimate the remaining service life, and the third task is to generate the optimal maintenance strategy based on equipment health assessment and maintenance cost constraints.

[0047] Step S5: When a pressure surge event is detected, start the fluid shock wave simulation, calculate the pressure peak value and arrival time at each node of the pipeline network, and generate the valve action sequence;

[0048] Step S6: Before implementing the control strategy, conduct a virtual rehearsal using a digital twin system to verify that there is no excessive negative pressure fluctuation before issuing the execution command.

[0049] Furthermore, in step S4:

[0050] A hard parameter sharing architecture is adopted, with the bottom convolutional layer extracting common features and the upper branch network processing different tasks respectively. The common features include first-order statistical features and second-order frequency domain features.

[0051] An uncertainty weighting mechanism is introduced to dynamically adjust the weighting coefficients based on the variance of the loss function of each task.

[0052] The model interpretation module is set up to generate a feature importance map through gradient backpropagation and label the top 5 key influencing factors.

[0053] Furthermore, in step S5:

[0054] The water hammer equation is solved using the method of characteristics to calculate the pressure wave propagation path and reflection superposition effect. The fluid parameters of the water hammer equation are derived from the unsteady fluid simulation model.

[0055] Establish a valve response delay model and set an action delay time of 0.5-2 seconds according to the actuator model;

[0056] Generate an instruction set containing an N+1 backup strategy, and automatically switch to the backup strategy when the primary strategy fails.

[0057] Compared with existing technologies, the intelligent fluid pipeline network management and control system and method based on multi-order spatiotemporal feature fusion described in this invention has the following advantages:

[0058] (1) The intelligent control system and method for fluid pipeline networks based on multi-level spatiotemporal feature fusion described in this invention significantly improves the control efficiency and safety of fluid pipeline networks through embedded multi-source sensing, multi-level spatiotemporal feature fusion and intelligent decision-making architecture:

[0059] (2) The fluid pipeline network intelligent control system and method based on multi-level spatiotemporal feature fusion described in this invention is based on 50kHz high-frequency sampling and encrypted transmission technology, combined with time domain statistics, frequency domain analysis and nonlinear feature extraction, which improves the accuracy of leak detection. Attached Figure Description

[0060] The accompanying drawings, which form part of this invention, are used to provide a further understanding of the invention. The illustrative embodiments of the invention and their descriptions are used to explain the invention and do not constitute an undue limitation of the invention. In the drawings:

[0061] Figure 1 This is a schematic diagram of the intelligent fluid pipeline network management and control system based on multi-order spatiotemporal feature fusion as described in an embodiment of the present invention;

[0062] Figure 2 This is a schematic diagram of the intelligent management and control method for fluid pipeline networks based on multi-level spatiotemporal feature fusion as described in an embodiment of the present invention. Detailed Implementation

[0063] It should be noted that, unless otherwise specified, the embodiments and features described in the present invention can be combined with each other.

[0064] In the description of this invention, it should be understood that the terms "center," "longitudinal," "lateral," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," and "outer," etc., indicating orientations or positional relationships based on the orientations or positional relationships shown in the accompanying drawings, are only for the convenience of describing the invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of the invention. Furthermore, the terms "first," "second," etc., are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Thus, a feature defined with "first," "second," etc., may explicitly or implicitly include one or more of that feature. In the description of this invention, unless otherwise stated, "a plurality of" means two or more.

[0065] In the description of this invention, it should be noted that, unless otherwise explicitly specified and limited, the terms "installation," "connection," and "linking" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal connection of two components. Those skilled in the art will understand the specific meaning of the above terms in this invention based on the specific circumstances.

[0066] The present invention will now be described in detail with reference to the accompanying drawings and embodiments.

[0067] This embodiment provides an intelligent management and control system for fluid pipeline networks based on multi-level spatiotemporal feature fusion, and the specific implementation is as follows:

[0068] Embedded IoT Sensing Layer Implementation: This layer is deployed at key nodes of the network. The core component includes an STM32H743 control motherboard that integrates a 4G / 5G dual-mode communication chip (model Quectel EC25). The motherboard has a built-in AES-256 encryption unit to ensure data transmission security.

[0069] The sensing unit consists of a triaxial vibration sensor array 7, a Coriolis flow meter 7, and a customized water intrusion sensor. The vibration sensor acquires signals at a sampling rate of 50kHz and performs seven-layer wavelet packet decomposition in real time through an ADSP-21489 chip. The water intrusion sensor obtains phase angle data by scanning in the 10-100kHz frequency band through an AD5933 impedance analyzer. All sensors are connected to the control motherboard via an RS-485 bus 7.

[0070] The actuator uses Bernard AQ series electric valves, which receive 4-20mA adjustment signals from the main board to achieve precise control of the valve opening.

[0071] Multi-order spatiotemporal feature processing layer implementation: This layer is deployed on edge computing nodes (NVIDIA Jetson AGXXavier). The spatiotemporal feature extraction module performs sliding window processing on pressure and flow velocity data, extracts the mean, variance and approximate entropy within the window to form first-order statistical features, calculates the first 32 harmonic amplitudes using 1024-point FFT as second-order frequency domain features, and calculates the Lyapunov exponent using the Rosenstein algorithm to form third-order nonlinear features.

[0072] The feature fusion module constructs an 8-head Transformer model, in which the MFCC features of the vibration signal are used as the key vector, the gray-level co-occurrence matrix of infrared thermal imaging is used as the value vector, and the cepstral coefficients of the acoustic signature are used as the query vector. The multi-head attention mechanism is used to achieve dynamic alignment of cross-modal features, and the attention weight matrix is ​​updated in real time at a frequency of 100Hz.

[0073] Intelligent decision-making layer implementation: The digital twin module constructs a three-dimensional pipeline model based on ANSYS Fluent, uses the finite volume method to solve the transient NS equations, and the dynamic boundary condition interface receives data from 300 pressure sensors at a frequency of 10Hz. The friction coefficient of the sealing ring is dynamically corrected according to the number of valve actions using the formula μ=0.15+0.02log(N).

[0074] The fault prediction module uses a residual LSTM network. The input layer has 128 nodes to receive spatiotemporal features. The residual connection spans three time steps and introduces a frequency domain feature branch. The training data contains 12% fault injection samples. The anomaly detection unit calculates the Mahalanobis distance between the real-time feature vector and the cluster center.

[0075] The linkage control module calculates the betweenness centrality of nodes based on the pipeline topology to classify valve priorities. Time-sensitive commands are transmitted through the RabbitMQ priority channel. The command retransmission mechanism triggers quantum key updates and command retransmission if no feedback is received within 2 seconds.

[0076] Cross-platform interaction layer implementation: The real-time in-memory database adopts a Redis cluster architecture, which is divided into a millisecond-level cache area, a feature intermediate cache area, and a persistent storage area for diagnostic results.

[0077] The network isolation unit uses a Huawei USG6350 firewall to assign the IoT sensing layer to an independent VLAN (ID 100), configuring a one-way communication policy that only allows outbound communication from the OPC UA port. The Modbus TCP parser filters write register commands, and a MAC whitelist binds six authorized device addresses. The visualization terminal renders a 3D pipeline model using a WebGL engine, employs HSL color mapping for the pressure cloud map, and synchronously displays the LSTM hidden layer activation status in the feature heatmap viewport. Abnormal areas are highlighted with flashing red lights.

[0078] This embodiment provides a method for intelligent management and control of fluid pipeline networks based on multi-order spatiotemporal feature fusion, and the specific implementation is as follows:

[0079] Step S1 Data Acquisition: Deploy embedded sensing terminals at key nodes of the pipeline network, integrating triaxial vibration sensors, wideband pressure transmitters, and distributed temperature sensor arrays. Vibration monitoring covers the characteristic frequency band of mechanical vibration, the pressure acquisition module is equipped with an adaptive range switching function, and the temperature sensor network adopts a bus topology to achieve continuous spatial monitoring. Multi-source data is acquired synchronously in time through an industrial-grade communication protocol. The raw signal is transmitted to the edge computing node after anti-aliasing filtering and format standardization. The transmission process adopts a data packet verification and retransmission mechanism to ensure integrity.

[0080] Step S2 Signal Reconstruction: To address the signal aliasing problem under complex pipeline network conditions, an improved empirical mode decomposition algorithm is used to separate the intrinsic modes of the original signal. Effective components are selected through frequency band energy analysis and time-frequency distribution correlation evaluation. The reconstructed feature signal retains the core information of fluid pulsation, mechanical vibration and heat conduction, while suppressing background noise interference. A sliding window mechanism is introduced in the reconstruction process to achieve dynamic adaptation of signal features, providing a high-quality input source for subsequent spatiotemporal feature extraction.

[0081] Step S3 Feature Cube Construction: Multi-dimensional operational features are extracted based on a sliding time window. The time-domain analysis module calculates waveform statistics and nonlinear dynamic indicators. The frequency-domain processing unit obtains energy distribution features through joint time-frequency analysis. The spatial correlation module calculates the parameter gradient change rate in combination with the pipeline network topology. Finally, the temporal evolution features, frequency-domain energy features, and spatial distribution features are organized into a three-dimensional data cube. This data structure fully represents the spatiotemporal evolution law of the pipeline network system and meets the input dimension requirements of the deep learning model.

[0082] Step S4 Multi-task learning: Construct a multi-task model with a parameter sharing mechanism. The bottom-level feature extraction network uses a temporal convolutional layer to capture cross-modal spatiotemporal correlations. The upper-level task branches respectively implement the functions of leak localization, lifetime prediction, and maintenance decision-making. During training, the learning progress of each task is balanced by adjusting the dynamic loss weights. The model interpretation module uses gradient backpropagation technology to identify key feature parameters, providing traceable technical basis for operation and maintenance decisions. In the model deployment stage, an online update mechanism is set up to realize the continuous optimization of the knowledge base.

[0083] Step S5 Pressure Sudden Change Response: When a pressure fluctuation exceeding the threshold event is detected, the fluid simulation engine solves the water hammer equation based on real-time data, predicts the pressure wave propagation path and intensity distribution, and generates a hierarchical control strategy by combining the pipeline topology and valve response characteristics. The main strategy prioritizes adjusting the near-end valves to suppress the wave propagation, while the backup strategy activates the compensation device to deal with extreme conditions. The strategy generation module simultaneously calculates the expected effects and assesses secondary risks, forming an optimized instruction set that balances safety and economy.

[0084] Step S6 Control Strategy Verification: Construct a multi-physics coupling verification platform in the digital twin environment. The fluid dynamics module verifies the pressure fluctuation suppression effect, the mechanical strength module evaluates the stress change of the pipe, and the thermo-coupling module analyzes the stability of the temperature field distribution. Through virtual pre-simulation, defective strategies that may cause system oscillation or equipment overload are eliminated. The verified instruction set is sent to the field execution unit through a dual-encryption channel. A status feedback closed-loop verification mechanism is set during the execution process to ensure the timing accuracy and reliability of the control actions.

[0085] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A fluid pipeline network intelligent control system based on multi-level spatiotemporal feature fusion, characterized in that, include: An embedded IoT sensing layer includes an electric valve control motherboard and / or a pneumatic valve control motherboard with an embedded remote information transmission module, as well as a flow sensor, a heat sensor and a water intrusion sensor connected to the control motherboard. Multi-level spatiotemporal feature processing layer, including: The spatiotemporal feature extraction module is configured to perform sliding window slicing on time-series data of pressure, flow velocity, and temperature parameters to extract first-order statistical features including mean and variance, second-order frequency domain features of Fourier transform coefficients, and third-order nonlinear features of Lyapunov exponents. The feature fusion module is configured to perform feature alignment on vibration spectrum, acoustic signature signal and infrared thermal imaging data through a multi-head attention mechanism; The intelligent decision-making layer, wherein the output of the multi-stage spatiotemporal feature processing layer is connected to the input of the intelligent decision-making layer through a message queue middleware, includes: The digital twin module constructs a visual model that includes the three-dimensional topology of the pipeline and fluid dynamics parameters; The fault prediction module includes an LSTM neural network with residual connections. The linkage control module is configured to generate an instruction queue containing valve identifiers and execution timing. The cross-platform interaction layer includes a Linux server running a real-time in-memory database and a Windows terminal that communicates with the server via the OPCUA protocol. The digital twin module: uses the finite volume method to construct an unsteady fluid simulation model, and the solver is configured to execute the SIMPLE algorithm; The dynamic boundary condition generation unit receives monitoring data from the pressure sensor and valve opening sensor in real time; Includes a database of corrected seal friction coefficients based on the number of valve opening and closing cycles; The linkage control module: Includes a valve priority sorting unit based on pipeline topology; Instruction queues are divided into time-sensitive queues and preventative maintenance queues; It is equipped with a command retransmission mechanism, and the retransmission trigger time threshold is 2 seconds. The 2-second threshold is determined based on a safety factor of 1.5 times the propagation speed of the pipeline pressure wave.

2. The intelligent control system for fluid pipeline networks based on multi-level spatiotemporal feature fusion according to claim 1, characterized in that, In the embedded IoT sensing layer: The remote information transmission module includes a 4G / 5G dual-mode communication chip and an AES-256 encryption unit; The vibration sensor array acquires vibration waveforms at a sampling rate of 50kHz and is equipped with a wavelet packet transform processor; The water intrusion sensor includes an impedance spectrum analysis circuit with an operating frequency range of 10-100kHz. The water intrusion sensor is connected to the control motherboard via an RS-485 bus to transmit the impedance spectrum phase angle change.

3. The intelligent control system for fluid pipeline networks based on multi-level spatiotemporal feature fusion according to claim 1, characterized in that, The fault prediction module: The training dataset contains samples labeled with normal operating conditions and 5%-15% fault injection conditions; The anomaly detection unit is configured to calculate the Mahalanobis distance between the real-time feature vector and the cluster center; The output layer includes an adaptive threshold adjustment unit based on ambient temperature; The residual connection is configured to span at least three time steps, and the skip links of the residual connection receive frequency domain feature input from the spatiotemporal feature extraction module.

4. The intelligent control system for fluid pipeline networks based on multi-level spatiotemporal feature fusion according to claim 2, characterized in that, The cross-platform interaction layer: The real-time in-memory database uses a Redis cluster architecture and includes millisecond-level data caching, feature intermediate result caching, and diagnostic result caching. The network isolation unit includes a hardware firewall and a MAC address-based whitelist access controller; The visualization terminal is equipped with a WebGL rendering engine and parallel pressure cloud viewport and feature heatmap viewport.

5. The intelligent control system for fluid pipeline networks based on multi-level spatiotemporal feature fusion according to claim 4, characterized in that, The network isolation unit: Divide the IoT sensing layer into independent security domains and configure unidirectional communication channels; Includes a Modbus TCP protocol parser, configured to filter write register instructions; The control command transmission channel is equipped with a quantum key distribution unit, which provides dynamic key updates for the AES-256 encryption unit.

6. A method for intelligent management and control of fluid pipeline networks based on multi-order spatiotemporal feature fusion, based on the system described in any one of claims 1-5, characterized in that, include: Step S1: Collect valve vibration spectrum, fluid pressure waveform, and pipe wall temperature gradient data through embedded sensing nodes, with a sampling frequency of not less than 10kHz. The valve vibration spectrum is collected through the vibration sensor array. Step S2: Using an improved empirical mode decomposition algorithm, the original signal is decomposed into intrinsic mode functions (IMFs), and the 2nd to 5th order IMF components are retained to reconstruct the characteristic signal. Step S3: Construct a spatiotemporal feature cube, divide 60 consecutive seconds of data into 20 overlapping time windows, and extract 128-dimensional time domain features and 64-dimensional frequency domain features from each window; Step S4: Train a multi-task learning model. The first task is to predict the location of the pipeline leak, the second task is to estimate the remaining service life, and the third task is to generate the optimal maintenance strategy based on equipment health assessment and maintenance cost constraints. Step S5: When a pressure surge event is detected, start the fluid shock wave simulation, calculate the pressure peak value and arrival time at each node of the pipeline network, and generate the valve action sequence; Step S6: Before implementing the control strategy, conduct a virtual rehearsal using a digital twin system to verify that there is no excessive negative pressure fluctuation before issuing the execution command.

7. The intelligent management and control method for fluid pipeline networks based on multi-order spatiotemporal feature fusion according to claim 6, characterized in that, In step S4: A hard parameter sharing architecture is adopted, with the bottom convolutional layer extracting common features and the upper branch network processing different tasks respectively. The common features include first-order statistical features and second-order frequency domain features. An uncertainty weighting mechanism is introduced to dynamically adjust the weighting coefficients based on the variance of the loss function of each task. The model interpretation module is set up to generate a feature importance map through gradient backpropagation and label the top 5 key influencing factors.

8. The intelligent management and control method for fluid pipeline networks based on multi-order spatiotemporal feature fusion according to claim 6, characterized in that, In step S5: The water hammer equation is solved using the method of characteristics to calculate the pressure wave propagation path and reflection superposition effect. The fluid parameters of the water hammer equation are derived from the unsteady fluid simulation model. Establish a valve response delay model and set an action delay time of 0.5-2 seconds according to the actuator model; Generate an instruction set containing an N+1 backup strategy, and automatically switch to the backup strategy when the primary strategy fails.

Images