Fracturing fleet power collaborative allocation and risk avoidance method

Abstract

The method for power collaborative distribution and risk avoidance of fracturing vehicle fleet relates to the technical field of fracturing vehicles, and is executed by a central collaborative controller which is in communication connection with local controllers corresponding to the fracturing vehicles; the method comprises the following steps: collecting running state parameters of each fracturing vehicle and receiving a total target of operation; constructing a dynamic collaborative state model of the vehicle fleet based on the running state parameters; establishing a multi-target collaborative optimization model; performing real-time solution on the multi-target collaborative optimization model to calculate target speed setting values; judging whether any target speed setting value is close to a preset safe working limit; judging whether there is a potential risk; selecting to adjust the target speed setting value according to the judgment result; and distributing final control instructions to the local controllers of the fracturing vehicles. The application can realize a global optimization control method with high robustness for power dynamic collaborative distribution of fracturing vehicle fleet, avoidance of running risks and the like.

Description

Technical Field

[0001] This invention relates to the field of fracturing truck technology, and in particular to a method for coordinated power allocation and risk avoidance in fracturing truck fleets. Background Technology

[0002] Fracturing trucks are core equipment for oil and gas field production enhancement operations, injecting high-pressure, high-volume fracturing fluid into wells to modify reservoirs. In large-scale operations, multiple fracturing trucks are often connected in parallel to form a pumping system to collectively meet the designed flow and pressure requirements. Traditional fleet control mainly relies on two methods: one is that the operator manually adjusts the speed of each truck according to the overall operation curve; the other is that a control system evenly distributes the total demand to each truck. Due to the complexity of operating conditions, factors such as formation pressure fluctuations and equipment differences continuously affect the output of individual fracturing trucks. The above-mentioned manual or static distribution-based models treat the fleet as a simple superposition of multiple independent units, lacking system-level dynamic coordination and information linkage. This results in slow control response, low accuracy, and difficulty in dealing with sudden changes, becoming a fundamental bottleneck restricting operational efficiency and safety.

[0003] In recent years, various technologies aimed at improving the performance of individual fracturing trucks have been developed. For example, algorithms such as reinforcement learning can be used to achieve adaptive speed control based on real-time status, improving the accuracy and response speed of a single machine. Other solutions focus on power system innovation, forming a hybrid power architecture by integrating energy storage, generators, and grid access, enabling a single fracturing truck to operate in multiple modes such as pure electric and hybrid. Furthermore, research has optimized the power matching between the engine and pump through reverse engineering or system verification to improve the economy and reliability of individual fracturing trucks. However, the optimization boundaries of these technologies are all limited to within a single fracturing truck. When multiple such vehicles form a fleet to operate collaboratively, each vehicle still operates independently according to its own logic, lacking state sharing and collaborative decision-making among vehicles. This results in the overall fleet efficiency potentially being lower than the sum of the capabilities of individual fracturing trucks, and even causing internal friction due to a lack of coordination.

[0004] Therefore, current technology faces the critical challenge of moving from "individual fracturing truck optimization" to "system optimization." Existing solutions have not yet addressed the following coordination issues: First, the dynamic load balancing problem. Due to differences in equipment status and efficiency, the actual working capacity boundaries of each vehicle in the fracturing fleet differ and change dynamically. Average distribution or independent control can easily lead to some vehicles operating at high loads for extended periods, while other vehicles are underloaded, triggering the "weakest link" effect, accelerating equipment wear and tear, and lacking a global load redistribution mechanism in the event of a single machine failure. Second, the systemic risk avoidance problem, especially the vibration risk of multi-machine coordination. When multiple vehicles are running simultaneously, the vibration frequencies of their engines and pumps may couple, inducing system resonance, a risk that exceeds the scope of single-machine control. Third, the fleet-level global energy efficiency optimization problem. Optimal energy consumption of a single fracturing truck does not equate to optimal overall fleet energy consumption. How to dynamically adjust the output of each vehicle to achieve the highest overall energy efficiency of the fleet while meeting total demand remains a technological gap. Therefore, there is an urgent need for a method that can achieve dynamic coordinated allocation of fleet power, intelligently avoid system risks, and optimize global energy efficiency to improve the overall efficiency, safety, and economy of fracturing operations. Summary of the Invention

[0005] In view of this, the technical problem to be solved by the present invention is to provide a method for power coordination and risk avoidance of fracturing vehicle fleet, which can realize dynamic power coordination and avoid systemic operational risks and has a highly robust global optimization control method.

[0006] To solve the above-mentioned technical problems, the technical solution of the present invention is as follows:

[0007] A method for coordinated power allocation and risk avoidance in fracturing truck fleets, wherein the method is executed by a central coordination controller, which is communicatively connected to the local controllers corresponding to each of the multiple fracturing trucks; the method includes the following steps:

[0008] S1. Real-time acquisition of operating status parameters for each fracturing truck and receipt of overall operational targets; operating status parameters include engine speed, output torque, fracturing pump outlet pressure, fracturing pump outlet flow rate, vibration spectrum characteristics, and equipment health status code;

[0009] S2. Based on the collected operating status parameters, construct a dynamic collaborative state model of the fleet. The dynamic collaborative state model is used to characterize the real-time capability boundary of each fracturing truck and the interaction between vehicles.

[0010] S3. Taking the achievement of the overall operation goal as the primary constraint, establish a multi-objective collaborative optimization model. The multi-objective collaborative optimization model takes minimizing the total energy consumption of the fleet, maximizing the load balance of the fleet, and minimizing the system resonance risk coefficient as optimization objectives.

[0011] S4. Solve the multi-objective collaborative optimization model in real time to calculate the target speed setting value of each fracturing truck in the next control cycle;

[0012] S5. Determine whether any calculated target speed setting value is close to the preset safe working limit of the corresponding fracturing truck;

[0013] S6. Based on the vibration spectrum characteristics of each fracturing truck, determine whether the fleet has a potential risk of inducing system resonance;

[0014] S7. Based on the judgment results of S5 and S6, select the corresponding risk avoidance strategy to adjust the target speed setpoint. The risk avoidance strategy includes load redistribution strategy and active frequency misalignment strategy, and generate the final control command based on the adjusted result.

[0015] S8. Distribute the final control command to the local controller of each fracturing truck, and each local controller controls its corresponding fracturing truck to execute the final control command.

[0016] Preferably, in S3, the multi-objective collaborative optimization model is implemented through a weighted comprehensive objective function. The weighted comprehensive objective function incorporates the resonance risk coefficient calculated in real time as a penalty term based on the integration of the fleet total energy consumption function and the load balance function.

[0017] Among them, the total energy consumption function of the fleet is calculated based on the universal characteristic curve of the engine of each fracturing truck and the current working condition; the load balance function is obtained by evaluating the deviation between the actual load and the average load of each truck; and the resonance risk coefficient is calculated in real time based on the frequency domain analysis of the vibration spectrum.

[0018] Preferably, the calculation steps for the resonance risk coefficient include:

[0019] Extract the main frequency and harmonic components of the real-time vibration spectrum of each fracturing truck; calculate the frequency difference f between the main frequencies of any two fracturing trucks; when the frequency difference f is less than a preset threshold, determine that there is a risk of frequency coupling, and calculate the resonance risk coefficient based on the degree of coupling and the magnitude of energy.

[0020] Preferably, in S7, the execution process of the load redistribution strategy includes:

[0021] When it is determined that the target speed setting of a fracturing truck exceeds the preset proportion of its safe working limit, the fracturing truck is marked as a load-limited object; based on the multi-objective collaborative optimization model, the target speed setting of other fracturing trucks is recalculated so that they can share the load increment of the load-limited object within the safe range, keeping the overall operation objective unchanged.

[0022] Preferably, in S7, the execution process of the active frequency misalignment strategy includes:

[0023] When multiple fracturing trucks are detected to have their main vibration frequencies coupled and the risk factor exceeds the threshold, select one or more fracturing trucks as the adjustment target; set a frequency offset for the adjustment target so that the frequency difference between its operating frequency and the main frequency of the other vehicles is greater than the safe frequency difference; recalculate the rotational speed setting value of the adjustment target.

[0024] Preferably, during the execution of S1 to S7, if it is determined by S5 that the target speed setting of a fracturing truck has exceeded its hard safety limit, or by S6 that the system resonance risk coefficient has exceeded the preset emergency threshold, the subsequent steps are immediately interrupted and an independent safety command is generated; the safety command includes an emergency stop command for a specific fracturing truck or a command to force the speed down to a safe speed; the central coordinating controller then executes S8, replacing any final control command to be sent with the safety command, and sends it directly to the corresponding local controller.

[0025] Preferably, it also includes a fault tolerance strategy; the execution process of the fault tolerance strategy includes:

[0026] During the execution of S1 to S7, the equipment health status codes and performance parameters of each fracturing truck are monitored in real time. If a fracturing truck is found to have a malfunction or a continuous abnormality in key performance parameters, it is immediately removed from the current collaborative fleet model. Based on the status of the remaining vehicles, execution restarts from S2 to generate new final control commands.

[0027] Preferably, the overall operational target includes the overall target displacement and the overall target pressure; the overall target displacement is calibrated by the wellhead flow meter, and the overall target pressure is determined based on the wellhead pressure sensor and the formation fracture pressure curve.

[0028] Preferably, the process of generating the device health status code includes:

[0029] Collect parameters such as engine cylinder pressure, lubricating oil pressure, transmission oil temperature, and bearing vibration energy; compare each parameter with preset thresholds, and generate a comprehensive status score based on the items and degrees exceeding the thresholds; map the score to a predefined equipment health status code.

[0030] Preferably, in S4, a model predictive control algorithm is used to solve the multi-objective collaborative optimization model in real time; the steps of the model predictive control algorithm include:

[0031] Using the current state as the initial condition, predict the changes in the fleet state within a finite time domain in the future; solve the optimization problem within this time domain to obtain the optimal control sequence; and output and execute the first control variable in the sequence, namely the target speed setpoint.

[0032] After adopting the above technical solution, the beneficial effects of the present invention are:

[0033] The fracturing truck fleet power coordination allocation and risk avoidance method of this application is executed by a central coordination controller, which communicates with the local controllers corresponding to each of the multiple fracturing trucks. The method includes the following steps: real-time acquisition of the operating status parameters of each fracturing truck and receiving the overall operational objective; the operating status parameters include engine speed, output torque, fracturing pump outlet pressure, fracturing pump outlet flow rate, vibration spectrum characteristics, and equipment health status code; based on the acquired operating status parameters, a dynamic coordination state model of the fleet is constructed, which is used to characterize the real-time capability boundary of each fracturing truck and the interaction between vehicles; with achieving the overall operational objective as the primary constraint, a multi-objective collaborative optimization model is established, which aims to minimize the total energy consumption of the fleet and the vehicle's overall energy consumption. The optimization objectives are to maximize the load balance of the fracturing trucks and minimize the system resonance risk coefficient. A multi-objective collaborative optimization model is solved in real time to calculate the target speed setpoint for each fracturing truck in the next control cycle. It is then determined whether any calculated target speed setpoint is close to the preset safe operating limit of the corresponding fracturing truck. Based on the vibration spectrum characteristics of each fracturing truck, it is assessed whether the fleet poses a potential risk of inducing system resonance. Based on the assessment results, corresponding risk avoidance strategies are selected to adjust the target speed setpoint. These risk avoidance strategies include load redistribution and active frequency shifting strategies. The final control command is generated based on the adjusted results. The final control command is then distributed to the local controllers of each fracturing truck, which in turn control their corresponding fracturing trucks to execute the final control command.

[0034] In the process of collaborative allocation and risk avoidance, this application firstly achieves a leap from independent control of single machines to fleet system optimization. Through global optimization and dynamic load balancing, it can significantly reduce the overall fuel or electricity consumption of the fleet while meeting the same construction requirements, thus improving operational economy. Secondly, through real-time vibration monitoring and resonance risk modeling, it can warn of and proactively avoid multi-machine resonance risks that are undetectable by traditional methods, enhancing operational safety and protecting expensive equipment. Furthermore, when a single vehicle fails or experiences a sudden performance drop, the system can automatically and quickly reassign tasks, achieving seamless degraded operation and ensuring the continuity and reliability of fracturing operations. In summary, this application not only improves the efficiency and economic benefits of fracturing operations, but more importantly, it constructs an intelligent risk warning and avoidance system at the system level, providing core technical support for achieving safe, efficient, and adaptive modern fracturing construction. Attached Figure Description

[0035] The present invention will be further described below with reference to the accompanying drawings and embodiments.

[0036] Figure 1 This is a flowchart of the fracturing vehicle fleet power coordination allocation and risk avoidance method according to an embodiment of the present invention. Detailed Implementation

[0037] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0038] like Figure 1 As shown, the fracturing truck fleet power coordination and risk avoidance method of the present invention is executed by a central coordination controller, which is communicatively connected to the local controllers corresponding to each of the multiple fracturing trucks; the method includes the following steps:

[0039] S1. Real-time acquisition of operating status parameters for each fracturing truck and receipt of overall operational targets; operating status parameters include engine speed, output torque, fracturing pump outlet pressure, fracturing pump outlet flow rate, vibration spectrum characteristics, and equipment health status code;

[0040] S2. Based on the collected operating status parameters, construct a dynamic collaborative state model of the fleet. The dynamic collaborative state model is used to characterize the real-time capability boundary of each fracturing truck and the interaction between vehicles.

[0041] S3. Taking the achievement of the overall operation goal as the primary constraint, establish a multi-objective collaborative optimization model. The multi-objective collaborative optimization model takes minimizing the total energy consumption of the fleet, maximizing the load balance of the fleet, and minimizing the system resonance risk coefficient as optimization objectives.

[0042] S4. Solve the multi-objective collaborative optimization model in real time to calculate the target speed setting value of each fracturing truck in the next control cycle;

[0043] S5. Determine whether any calculated target speed setting value is close to the preset safe working limit of the corresponding fracturing truck;

[0044] S6. Based on the vibration spectrum characteristics of each fracturing truck, determine whether the fleet has a potential risk of inducing system resonance;

[0045] S7. Based on the judgment results of S5 and S6, select the corresponding risk avoidance strategy to adjust the target speed setpoint. The risk avoidance strategy includes load redistribution strategy and active frequency misalignment strategy, and generate the final control command based on the adjusted result.

[0046] S8. Distribute the final control command to the local controller of each fracturing truck, and each local controller controls its corresponding fracturing truck to execute the final control command.

[0047] In this application, the central coordinating controller does not simply command each fracturing truck individually, but treats the entire fleet as a dynamically controllable organic whole. First, it acquires two types of key information in real time via S1: one is comprehensive operational status parameters, including mechanical vibration spectrum and equipment health status codes; the other is the required total displacement and total pressure. In S2, the central coordinating controller uses these real-time operational status parameters to construct a dynamic coordinating state model of the fleet. This model not only quantifies the upper limit of each truck's capabilities at the current moment (such as the maximum safe speed based on health status), but also estimates the mutual mechanical influence between vehicles through shared pipelines and infrastructure, laying a precise digital foundation for global decision-making.

[0048] Based on the dynamic collaborative state model, a multi-objective optimization problem is established in S3. Its primary task is to meet the total displacement and total pressure required for construction. Under this rigid constraint, three optimization objectives are pursued simultaneously: reducing the fuel or energy consumption of the entire fracturing truck fleet, balancing the load among the trucks, and minimizing the risk of system resonance caused by multi-machine vibration coupling. S4 then solves this complex optimization problem quickly in real time, ultimately outputting a set of personalized target speed settings for each fracturing truck. S5 and S6 perform a dual safety check on this preliminary solution: first, checking whether any vehicle has been assigned a speed task exceeding its own safe operating limit; second, judging from the vibration spectrum characteristics whether there is a potential risk of overall structural resonance if the fleet operates at this speed.

[0049] After safety screening, in S7, the central coordination controller intelligently selects and triggers corresponding risk avoidance strategies. If a vehicle's load is too high, a load redistribution strategy is activated, automatically transferring part of its load to other vehicles with sufficient capacity while ensuring the overall objective remains unchanged. If resonance risk is detected, an active frequency shifting strategy is activated, fine-tuning the speed of relevant vehicles to avoid dangerous frequency bands. Through these fine adjustments, the final, safe, and optimized coordination control commands are generated. Finally, in S8, these commands are distributed to the local controllers of each vehicle for execution, thus translating the central optimization decision into the actual actions of the fleet. The entire process restarts in the next control cycle (usually within seconds), forming a continuous closed loop of perception-modeling-optimization-verification-adjustment-execution, achieving real-time tracking and dynamic optimal control of the fleet status.

[0050] First, this application represents a paradigm shift from independent single-machine control to fleet system optimization. Through global optimization and dynamic load balancing, it significantly reduces overall fleet fuel or electricity consumption while meeting the same operational requirements, thereby improving operational economy. Second, through real-time vibration monitoring and resonance risk modeling, this application can proactively mitigate multi-machine resonance risks that are undetectable by traditional methods, greatly enhancing operational safety and protecting expensive equipment. Third, the fault tolerance and emergency response mechanisms significantly improve system robustness. When a single vehicle fails or experiences a sudden performance drop, the system can automatically and quickly redistribute tasks, achieving seamless degraded operation and ensuring the continuity and reliability of fracturing operations. In summary, this application not only improves the efficiency and economic benefits of fracturing operations but, more importantly, constructs an intelligent risk warning and avoidance system at the system level, providing core technical support for achieving safe, efficient, and adaptive modern fracturing operations.

[0051] In this application, the real-time acquisition of the operating status parameters of each fracturing truck in S1, along with the receipt of the overall operational target, serves as the data sensing and input starting point for achieving closed-loop collaborative control. Its core lies in the simultaneous acquisition of two types of information, providing a foundation for all subsequent analysis, decision-making, and control.

[0052] The first type of information consists of refined operating status parameters for each fracturing truck, a multi-dimensional dataset. Specifically, it includes: engine speed and output torque, which characterize the instantaneous output capacity and operating conditions of the power system, serving as a direct basis for calculating power and energy consumption; fracturing pump outlet pressure and flow rate, which directly reflect the effectiveness of fracturing operations and are core indicators for assessing whether process requirements are met; vibration spectrum characteristics, obtained through frequency domain analysis of mechanical vibration signals, revealing the operating frequency, harmonic components, and energy distribution of rotating components, providing key dynamic characteristics for assessing mechanical condition and predicting system resonance risk; and equipment health status codes, which are pre-processed comprehensive assessment results based on the analysis of multi-source sensor data, quantifying the overall health and performance degradation of the equipment in coded form, directly defining the reliable operating boundary.

[0053] The second type of information is the overall operational objective, which comes from the upper-level design and specifies the final flow and pressure requirements that the entire fleet needs to achieve in its coordinated operations. It serves as the overall constraint and guide for all optimization and control.

[0054] The acquisition of operating status parameters is accomplished through the local controller installed on each fracturing truck and its connected sensor network. Engine speed and output torque are directly read from the engine electronic control unit via the controller's local area network bus. The fracturing pump outlet pressure and flow rate are measured by pressure transmitters and flow meters installed on the high-pressure manifold and the signals are transmitted to the local controller. Vibration spectrum characteristics are acquired by vibration acceleration sensors, converted by a data acquisition card, and then extracted in real time by the local controller using a fast Fourier transform algorithm. The equipment health status code is not a raw signal, but a comprehensive evaluation code generated online by the local controller by comprehensively analyzing multi-source monitoring parameters such as engine cylinder pressure and lubricating oil temperature and combining them with threshold rules. All these acquired or generated parameters are uploaded by each local controller to the central collaborative controller via industrial Ethernet or wireless private network according to a preset protocol cycle.

[0055] Meanwhile, receiving the overall operational target involves interaction between the central coordinating controller and the upper-level system, primarily through two methods: first, the operator manually inputs or confirms the target values ​​for total displacement and total pressure on the human-machine interface in the control room based on the construction design; second, the central coordinating controller automatically receives instructions issued in real-time by the upper-level fracturing construction optimization and command system based on the geological engineering model via a standard data interface. The overall operational target includes the target displacement and the target pressure; the target displacement is calibrated using a wellhead flow meter, and the target pressure is determined based on the wellhead pressure sensor and the formation fracture pressure curve.

[0056] In summary, the acquisition of operational status parameters is a bottom-up distributed sensing process, while the reception of the overall operational objective is a top-down command issuance process. Both converge to the central collaborative controller through a reliable industrial communication network, together forming the information input foundation for the entire collaborative optimization and control method.

[0057] In S2, constructing a dynamic collaborative state model for the fracturing fleet refers to establishing a digital model to describe the dynamic characteristics of the fleet system using collected operational state parameters. Through data processing and system identification, the model calculates and updates two core elements in real time: the real-time capability boundary of each fracturing truck, and the interaction effects between vehicles.

[0058] Specifically, in S2, constructing a dynamic cooperative state model includes the following steps:

[0059] S21. Construct a single-machine dynamic capability boundary model based on health status;

[0060] For each fracturing truck, based on its equipment health status code and real-time operating conditions, determine its maximum permissible power output range within the current control cycle; including:

[0061] The central coordinating controller reads the device health status code uploaded by each local controller. ( Indicates the first (A fracturing truck), and obtains the corresponding reduction coefficient based on a preset mapping table between health status codes and reduction coefficients. ;

[0062] Based on the universal characteristic curve data of the fracturing truck engine (including rated speed) With rated torque In addition to real-time data on transmission oil temperature and lubricating oil pressure, the maximum permissible engine speed of the vehicle is calculated. and maximum torque The formula is as follows:

[0063] =

[0064] =

[0065] in, and This is a condition correction factor based on real-time oil temperature and oil pressure;

[0066] Determine the current maximum allowable output power :

[0067] =

[0068] in, This is a comprehensive efficiency correction factor based on the current operating conditions, used to dynamically derate the theoretical maximum power;

[0069] Output the set of allowed working domains for each fracturing truck. , serving as the real-time capability boundary in the dynamic collaborative state model;

[0070] S22. Construct a hydraulic interaction influence matrix for the vehicle fleet;

[0071] Define hydraulic sensitivity coefficient , indicating the first The effect of changes in fracturing truck flow rate on the first The degree of influence of the trolley pump inlet pressure is calculated using the following formula:

[0072]

[0073] Real-time data collection of vehicle exit flow Construct a hydraulic interaction matrix using system pressure data. This reflects the aforementioned coupling relationship; It is a dynamic pipeline impedance matrix, whose value is updated in real time through online identification as the operating conditions change, so as to characterize the nonlinear fluid resistance characteristics under high pressure.

[0074] S23. Construct a mechanical vibration coupling transmission model;

[0075] Based on the vibration spectrum characteristics collected by S1, the dominant vibration frequency of each fracturing truck is identified. and amplitude This includes: peak detection of the vibration spectrum, and determining the frequency corresponding to the peak value with the highest energy or amplitude as the vehicle's main frequency. ; will increase the main frequency The corresponding spectral amplitude is used as the amplitude. ;

[0076] A vibration attenuation transfer function is established using pre-determined site transfer characteristic parameters and equipment spacing. , describing the The vibration of the trolley is transmitted to the first Energy attenuation and phase difference at different trolley positions; calculation of potential resonance superposition energy of the convoy. The formula is as follows:

[0077]

[0078] in, This is a function to determine the degree of frequency similarity. Used to quantify the superposition state of vibration energy when multiple vehicles have similar frequencies, forming the mechanical interaction influence part in the dynamic cooperative state model;

[0079] S24. Generate the integrated state-space equations;

[0080] The above set of single-machine capability boundaries Hydraulic interaction matrix Coupled with vibration model (with) (Representation) integration forms a unified dynamic cooperative state model, whose state-space form is:

[0081]

[0082] in, This is the system state vector, which includes the speed, pressure, vibration status, etc. of each vehicle; The target speed control vector; This is the set of boundary constraints for the capabilities of each vehicle; this model is used as the predictive model for subsequent model predictive control (MPC). Specifically, in S4, the model predictive control algorithm is used to solve the multi-objective cooperative optimization model in real time. The steps of this algorithm are: using the current state as the initial condition, using the predictive model (i.e., the dynamic cooperative state model constructed in S2) to predict the future state and solve the optimization, and calculating the target speed setpoint for each fracturing truck in the next control cycle.

[0083] Determining the real-time capability boundary relies on dynamically input device health status codes and current operating conditions. For example, a health status code indicating performance degradation will cause the model to lower the maximum permissible torque and speed thresholds of the vehicle's engine. Simultaneously, by combining real-time parameters such as current engine speed and output torque, the model can assess its thermal and mechanical loads, thereby dynamically calculating the upper limit of power and speed that the vehicle can safely output in the next cycle. This makes the capability boundary a dynamic value that changes with the state.

[0084] The interactions between vehicles primarily involve mechanical vibration coupling and hydraulic system coupling. For mechanical coupling, the model analyzes the proximity and energy of the dominant frequencies of each fracturing truck based on their vibration spectrum characteristics, and calculates the transmission and superposition effects of vibration energy within the fleet structure, taking into account the vehicle's physical layout. For hydraulic coupling, the model uses fluid dynamics principles, based on the pump outlet pressure and flow rate of each fracturing truck and the shared manifold network topology, to estimate the back pressure disturbance caused to other vehicles when the output of one fracturing truck changes. Ultimately, the model outputs a system state description containing the capability boundaries and interaction matrix of all vehicles, providing precise system-level constraints for subsequent global optimization, enabling the central controller to perform overall coordinated control of the fleet.

[0085] In S3, the multi-objective collaborative optimization model is achieved through a weighted comprehensive objective function. The weighted comprehensive objective function integrates the total energy consumption function of the fleet and the load balance function, and incorporates the resonance risk coefficient calculated in real time as a penalty term. The total energy consumption function of the fleet is calculated based on the universal characteristic curve of the engine of each fracturing truck and the current operating conditions. The load balance function is obtained by evaluating the deviation between the actual load and the average load of each truck. The resonance risk coefficient is calculated in real time based on the frequency domain analysis of the vibration spectrum.

[0086] The multi-objective collaborative optimization model is implemented through a weighted comprehensive objective function. First, it integrates two objective functions to be minimized: the total fleet energy consumption function and the fleet load balance function. Second, it incorporates a real-time calculated resonance risk coefficient as a penalty term. The total fleet energy consumption function is calculated based on the universal characteristic curves of each fracturing truck's engine and its real-time operating conditions to quantify fuel consumption at different speeds and torques. The load balance function is obtained by evaluating the deviation between the actual load of each truck and the average load of the fleet, aiming to make the load of each truck as uniform as possible. The resonance risk coefficient is calculated based on frequency domain information analyzed in real-time from vibration spectrum characteristics; it quantifies the probability and severity of resonance caused by multi-truck vibration frequency coupling. Through this design, the weighted comprehensive objective function guides the optimization solution process, ensuring that the overall operational objective is met while finding a fleet control strategy that achieves the optimal balance between reducing total fuel consumption, balancing the load of each truck, and avoiding system resonance risk.

[0087] The steps for calculating the resonance risk coefficient include:

[0088] Extract the main frequency and harmonic components of the real-time vibration spectrum of each fracturing truck; calculate the frequency difference f between the main frequencies of any two fracturing trucks; when the frequency difference f is less than a preset threshold, determine that there is a risk of frequency coupling, and calculate the resonance risk coefficient based on the degree of coupling and the magnitude of energy.

[0089] In S7, the execution process of the load redistribution strategy includes: when it is determined that the target speed setting of a fracturing truck exceeds the preset proportion of its safe working limit, the fracturing truck is marked as a load-limited object; based on the multi-objective collaborative optimization model, the target speed setting of other fracturing trucks is recalculated so that they can share the load increment of the load-limited object within the safe range, keeping the overall operation objective unchanged.

[0090] When the central coordinating controller determines that the target speed setpoint of a fracturing truck exceeds a preset proportion of its safe operating limit, the truck will be marked as a load-limited vehicle. Subsequently, the central coordinating controller recalculates the target speed setpoints of the remaining fracturing trucks in the fleet based on the original multi-objective collaborative optimization model. The goal of the recalculation is to ensure that these vehicles, within their respective safe operating ranges, share the portion of the load increment originally borne by the load-limited vehicle, thereby maintaining the overall total displacement and total pressure targets unchanged. This process dynamically adjusts the output of each vehicle through optimization algorithms, achieving a safe transfer of load from overloaded vehicles to vehicles with surplus capacity, ensuring the continuity and safety of the fleet's collaborative operations.

[0091] In S7, the execution process of the active frequency misalignment strategy includes: when multiple fracturing trucks are detected to have vibration main frequency coupling and the risk factor exceeds the threshold, one or more fracturing trucks are selected as the adjustment object; a frequency offset is set for the adjustment object so that the frequency of the adjusted object is greater than the main frequency difference of the other vehicles; and the speed setting value of the adjustment object is recalculated.

[0092] When multiple fracturing trucks are detected to have coupled vibration frequencies and the calculated resonance risk coefficient exceeds a preset threshold, the central coordination controller selects one or more of these trucks as adjustment targets. Then, a specific frequency offset is set for the selected target, ensuring its adjusted operating frequency maintains a distance greater than the safe frequency difference from the main frequencies of the remaining trucks in the fleet. Finally, the target rotational speed setting is recalculated based on this frequency offset. By actively and purposefully adjusting the operating frequencies of potential resonance source vehicles, the conditions for multi-machine frequency coupling are fundamentally disrupted, thus preventing systemic resonance at the vibration dynamics level. This improves the system's dynamic stability in high-power-density, multi-machine parallel operation scenarios, avoiding structural fatigue damage, sealing failures, and even cascading equipment failures that resonance may cause, ensuring the safety and reliability of fracturing operations. Simultaneously, the precise recalculation of rotational speed ensures that core operational indicators such as displacement and pressure remain unaffected.

[0093] It is worth noting that simply identifying frequency coupling is insufficient to determine the level of resonance risk; a quantitative assessment is also necessary. The main criteria are: first, the tightness of the coupling—the smaller the frequency difference *f*, the tighter the coupling and the higher the baseline risk; second, the magnitude of the vibrational energy involved in the coupling—the larger the amplitude corresponding to the dominant vibration frequency, the greater the excitation energy it contains, and the more severe the potential resonance response. The frequency difference and vibrational energy are combined using a predefined model or algorithm to calculate the specific system resonance risk coefficient under the current state. Finally, the system resonance risk coefficient is compared with a preset safety threshold and a higher emergency threshold to determine the risk level, which includes no risk, potential risk (requiring monitoring or adjustment), or emergency risk (requiring immediate interruption).

[0094] This application also includes an emergency protection mechanism. During the execution of S1 to S7, if it is determined by S5 that the target speed setting of a fracturing truck has exceeded its hard safety limit, or by S6 that the system resonance risk coefficient has exceeded the preset emergency threshold, the subsequent steps will be immediately interrupted and an independent safety command will be generated. The safety command includes an emergency shutdown command for a specific fracturing truck or a command to force the speed to be reduced to a safe speed. The central coordinating controller then executes S8, replacing any final control command to be sent with the safety command, and sends it directly to the corresponding local controller.

[0095] When a fracturing truck is identified in the collaborative optimization process as being about to exceed an insurmountable hard safety limit, or when the system faces a sudden surge in resonance risk, the emergency protection mechanism forcibly exits all normal optimization calculations and strategy adjustment cycles. This completely avoids protection delays that may occur due to continuous calculations or strategy iterations in emergency situations. By generating a completely independent and clearly defined safety instruction (such as emergency shutdown or forced deceleration), and having the central collaborative controller directly execute the instruction distribution step, the mechanism ensures that in critical moments, the control instruction flow is concise, unique, and clearly targeted. The rule of replacing any pending final control instructions with safety instructions eliminates any possibility of confusion or competition between safety instructions and potentially risky optimization instructions. This forms a clear safety hierarchy with load optimization and proactive frequency avoidance: the former optimizes performance within safety boundaries, while the emergency protection mechanism is the ultimate protection measure when safety boundaries are about to be breached or a systemic crisis has been detected. Therefore, as the safety bottom line guarantee for the entire collaborative control system, the emergency protection mechanism enhances the system's survivability and reliability under extreme conditions or sudden anomalies, ensuring a closed loop in risk management.

[0096] This application also includes a fault tolerance strategy; the execution process of the fault tolerance strategy includes: during the execution of S1 to S7, real-time monitoring of the equipment health status code and performance parameters of each fracturing truck; when a fracturing truck is found to have a fault or continuous abnormality in key performance parameters, it is immediately removed from the current collaborative fleet model, and based on the status of the remaining vehicles, execution is restarted from S2 to generate new final control instructions.

[0097] The fault tolerance strategy of this application executes an independent real-time monitoring task in parallel during the continuous operation of S1 to S7. The monitoring objects are the equipment health status codes and key performance parameters of each fracturing truck.

[0098] When the monitoring logic determines that a fracturing truck has malfunctioned or its key performance parameters show continuous anomalies, indicating that the truck can no longer reliably perform the predetermined collaborative tasks, the strategy is immediately activated. Its core operation is to remove the malfunctioning vehicle from the dynamic collaborative state model representing the overall fleet status. This means that subsequent optimization calculations will no longer consider any capability contribution or interaction impact of that vehicle. Subsequently, the system uses the remaining, normally functioning vehicles as the target and restarts the complete collaborative optimization and control command generation process from S2.

[0099] Fault-tolerant strategies endow the entire collaborative control system with online adaptive reconfiguration and degraded operation capabilities. First, the fault-tolerant strategy achieves proactive fault isolation and system reconfiguration. By removing faulty units from the control model in real time, it prevents their abnormal states from contaminating the global optimization results or sending erroneous commands to other vehicles, ensuring the integrity of the core control logic. Second, it ensures operational continuity. In the event of partial equipment failure, the system does not collapse or simply issue an alarm. Instead, it automatically and quickly recalculates a feasible and optimized collaborative solution based on remaining resources, enabling the fleet to continue operations with reduced overall capacity and minimizing unplanned downtime. Finally, it enhances the overall availability and robustness of the system, allowing it to cope with unavoidable equipment emergencies in actual operations, localizing the impact of single-point failures and achieving a smooth degradation of system functionality. Therefore, fault-tolerant strategies are not an independent remedial measure, but rather a key resilience design integrated into the main collaborative optimization process, ensuring the long-term reliable operation of the system.

[0100] The process of generating equipment health status codes includes: collecting parameters such as engine cylinder pressure, lubricating oil pressure, transmission oil temperature, and bearing vibration energy; comparing each parameter with preset thresholds, generating a comprehensive status score based on the items and degrees exceeding the thresholds; and mapping the score to a predefined equipment health status code.

[0101] First, key parameters such as engine cylinder pressure, lubricating oil pressure, transmission oil temperature, and bearing vibration energy are collected. Then, the measured values ​​of each parameter are compared with preset thresholds. Based on the parameters exceeding limits, the magnitude of the exceedance, and the duration, the system calculates a comprehensive status score according to predetermined rules. Finally, this score is converted into a standardized equipment health status code through a preset mapping relationship (e.g., using numbers 0 to 3 to represent healthy, warning, performance degradation, and serious fault, respectively). This status code provides the central coordinating controller with a clear quantitative indicator of equipment health, directly used to dynamically define the vehicle's safe operating capability boundaries, and provides real-time basis for fault warning and system fault-tolerant decision-making.

[0102] In S4, a model predictive control algorithm is used to solve the multi-objective cooperative optimization model in real time. The steps of the model predictive control algorithm include: using the current state as the initial condition, predicting the changes in the fleet state within a finite time domain in the future; solving the optimization problem within this time domain to obtain the optimal control sequence; and outputting the first control variable in the sequence, namely the target speed setpoint, for execution.

[0103] The model predictive control algorithm operates based on the principles of rolling optimization and feedback correction. First, using the current dynamic coordination state of the fleet as initial conditions, it predicts the system state changes over a finite time domain. Then, within this prediction time domain, it solves a multi-objective optimization problem to obtain an optimal control sequence. Finally, it outputs only the first control variable in this sequence, i.e., the target speed setpoint, for execution. After each control cycle, the system updates the initial conditions based on the latest measured state and repeats the above process.

[0104] The optimization process takes into account the subsequent effects of control actions, thereby achieving smoother and more rational collaborative decision-making while satisfying multiple objectives. The rolling optimization mechanism, combined with real-time state feedback, enables the system to continuously adapt to model errors and external disturbances, ensuring the continuous effectiveness and stable execution of the collaborative control strategy in complex dynamic environments.

[0105] 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, and improvements made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for power coordination and risk avoidance in a fracturing fleet, characterized in that, The method is executed by a central collaborative controller, which is communicatively connected with local controllers corresponding to the plurality of fracturing trucks respectively; the method comprises the following steps: S1, collecting running state parameters of each fracturing truck in real time, and receiving a total job target; the running state parameters include engine speed, output torque, fracturing pump outlet pressure, fracturing pump outlet flow, vibration frequency spectrum characteristics and device health status code; S2, constructing a dynamic collaborative state model of the truck fleet based on the collected running state parameters, the dynamic collaborative state model being used to represent real-time capability boundaries of each fracturing truck and interaction influences between the vehicles; the central collaborative controller constructs the dynamic collaborative state model of the truck fleet by using the real-time running state parameters, quantifies the upper limit of the capability of each truck at the current time, and estimates the mutual mechanical influences between the vehicles through shared pipelines and infrastructure; S3, establishing a multi-objective collaborative optimization model with the total job target as the primary constraint condition, the multi-objective collaborative optimization model taking minimization of total energy consumption of the truck fleet, maximization of load balance degree of the truck fleet and minimization of system resonance risk coefficient as optimization objectives; the multi-objective collaborative optimization model is realized through a weighted comprehensive objective function, which integrates the total energy consumption function of the truck fleet and the load balance degree function, and incorporates the real-time calculated resonance risk coefficient as a penalty term; wherein, the total energy consumption function of the truck fleet is calculated according to the engine universal characteristic curve of each fracturing truck and the current working condition; the load balance degree function is obtained by evaluating the deviation of the actual load of each truck from the average load; the resonance risk coefficient is calculated in real time based on frequency domain analysis of the vibration frequency spectrum; S4, solving the multi-objective collaborative optimization model in real time to calculate the target speed set value of each fracturing truck in the next control period; the model predictive control algorithm is used to solve the multi-objective collaborative optimization model in real time; the steps of the model predictive control algorithm include: taking the current state as the initial condition, predicting the state change of the truck fleet in a finite time domain in the future; solving the optimization problem in the time domain to obtain the optimal control sequence; outputting and executing the first control quantity in the sequence, i.e. the target speed set value; S5, judging whether any target speed set value calculated is close to the preset safe working limit of the corresponding fracturing truck; S6, judging whether there is a potential risk of inducing system resonance in the truck fleet based on the vibration frequency spectrum characteristics of each fracturing truck; S7, selecting a corresponding risk avoidance strategy to adjust the target speed set value according to the judgment results of S5 and S6, the risk avoidance strategy including a load redistribution strategy and an active frequency offset strategy, and generating a final control instruction based on the adjusted result; S8, distributing the final control instruction to the local controllers of the fracturing trucks, and controlling the corresponding fracturing trucks to execute the final control instruction by the local controllers.

2. The frac fleet power co-assignment and risk aversion method of claim 1, wherein, The calculation steps of the resonance risk coefficient include: extracting the main frequency and harmonic components of the real-time vibration frequency spectrum of each fracturing truck; calculating the frequency difference f between the vibration main frequencies of any two fracturing trucks; when the frequency difference f is less than a preset threshold, it is determined that there is a frequency coupling risk, and the resonance risk coefficient is calculated according to the coupling degree and energy size.

3. The frac fleet power co-assignment and risk aversion method of claim 1, wherein, In S7, the execution process of the load redistribution strategy includes: When it is determined that the target speed set value of a fracturing truck exceeds the preset proportion of its safe working limit, the fracturing truck is marked as a load-limited object; based on the multi-objective collaborative optimization model, the target speed set values of other fracturing trucks are recalculated to share the load increment of the load-limited object within the safe range, and the total target is kept unchanged.

4. The frac fleet power co-assignment and risk aversion method of claim 1, wherein, In S7, the execution process of the active mistuning strategy includes: When it is detected that the vibration main frequencies of multiple fracturing trucks are coupled and the risk coefficient exceeds the threshold value, one or more of the fracturing trucks are selected as adjustment objects; a frequency offset is set for the adjustment objects, so that the operating frequency of the adjustment objects after adjustment is separated from the main frequencies of the remaining vehicles by more than the safe frequency difference; the speed set value of the adjustment object is recalculated.

5. The frac fleet power co-assignment and risk aversion method of claim 1, wherein, In the execution process of S1 to S7, if it is determined through S5 that the target speed set value of a fracturing truck has exceeded its hard safety limit, or it is determined through S6 that the system resonance risk coefficient exceeds the preset emergency threshold value, the subsequent steps are immediately interrupted, and an independent safety instruction is generated; the safety instruction includes an emergency shutdown instruction or a forced speed reduction instruction to a safe speed for a specific fracturing truck; the central collaborative controller immediately executes S8 to replace any final control instruction to be sent with the safety instruction and sends it directly to the corresponding local controller.

6. The frac fleet power co-assignment and risk aversion method of claim 1, wherein, It also includes a fault tolerance strategy; The execution process of the fault tolerance strategy includes: In the execution process of S1 to S7, the device health status code and performance parameters of each fracturing truck are monitored in real time; when it is detected that a fracturing truck has a fault or a continuous abnormality in a key performance parameter, it is immediately excluded from the current collaborative vehicle model, and based on the state of the remaining vehicles, the execution is restarted from S2 to generate new final control instructions.

7. The frac fleet power co-assignment and risk aversion method of claim 1, wherein, The total target includes the total target displacement and the total target pressure; the total target displacement is calibrated by the wellhead flowmeter, and the total target pressure is determined according to the wellhead pressure sensor and the formation breakdown pressure curve.

8. The frac fleet power co-assignment and risk aversion method of claim 1, wherein, The generation process of the device health status code includes: Engine cylinder pressure, lubricating oil pressure, transmission oil temperature, bearing vibration energy parameters are collected; each parameter is compared with the preset threshold value, and a comprehensive state score is generated according to the items and degrees exceeding the threshold value; the score is mapped to a predefined device health status code.

Images