Fracturing fleet power collaborative allocation and risk avoidance method

Abstract

A fracturing truck fleet power collaborative distribution and risk avoidance method relates to the technical field of fracturing trucks, and is executed by a central collaborative controller, and the central collaborative controller is in communication connection with local controllers corresponding to a plurality of fracturing trucks; the method comprises the following steps: collecting operation state parameters of each fracturing truck, and receiving a total operation target; based on the operation state parameters, constructing a dynamic collaborative state model of the motorcade; establishing a multi-objective collaborative optimization model; solving the multi-target collaborative optimization model in real time, and calculating a target rotating speed set value; judging whether any target rotating speed set value is close to a preset safety working limit or not; judging whether a potential risk exists or not; according to the judgment result, the target rotating speed set value is selected to be adjusted; and distributing the final control instruction to a local controller of each fracturing truck. According to the method, dynamic collaborative distribution of fracturing fleet power can be realized, operation risks are avoided, and the method is a global optimization control method with high robustness.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of fracturing trucks, and particularly relates to a method for power coordination and risk avoidance of a fleet of fracturing trucks. BACKGROUND

[0002] A fracturing truck is a core equipment for oil and gas field stimulation, which injects high-pressure and high-flow-rate fracturing fluid into a well to transform a reservoir. In large-scale construction, multiple fracturing trucks are often connected in parallel to form a pumping system to meet the design requirements of flow rate and pressure. Traditional truck fleet control mainly relies on two methods: one is to manually adjust the speed of each truck according to the total construction curve by an operator; and the other is to use a control system to evenly distribute the total demand to each truck. Due to the complex working conditions, factors such as fluctuation of formation pressure and equipment differences continuously affect the output of a single fracturing truck. The above-mentioned manual or static distribution mode treats the truck fleet as a simple superposition of multiple independent units, lacks dynamic coordination and information linkage at the system level, and leads to slow control response, low precision, and difficulty in responding to sudden changes, which becomes a basic bottleneck restricting the efficiency and safety of the operation.

[0003] In recent years, various technologies aimed at improving the performance of a single fracturing truck have been developed. For example, using reinforcement learning algorithms can achieve adaptive speed control based on real-time state, improving the accuracy and response speed of a single machine. Another scheme focuses on power system innovation, integrating energy storage, generators, and grid access to form a hybrid power architecture, allowing a single fracturing truck to operate in pure electric, hybrid, and other modes. In addition, there are studies that optimize the power matching of engines and pumps through reverse calculation or system checking to improve the economy and reliability of a single fracturing truck. However, the optimization boundaries of these technologies are limited to the internal of a single fracturing truck. When multiple such vehicles form a fleet for cooperative operation, each vehicle still operates independently according to its own logic, and there is a lack of state sharing and cooperative decision-making between vehicles. This leads to the overall efficiency of the fleet being lower than the sum of the capabilities of individual fracturing trucks, and even generates internal consumption due to the 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, real-time solving the multi-objective collaborative optimization model to obtain the target speed set value of each fracturing truck in the next control period;

[0012] S5, judging whether any target speed set value obtained is close to the preset safe working limit of the corresponding fracturing truck;

[0013] S6, judging whether the vehicle team has a potential risk of inducing system resonance based on the vibration spectrum characteristics of each fracturing truck;

[0014] S7, according to the judgment results of S5 and S6, selecting a corresponding risk avoidance strategy to adjust the target speed set value, 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;

[0015] S8, distributing the final control instruction to the local controller of each fracturing truck, and controlling the corresponding fracturing truck to execute the final control instruction by each local controller.

[0016] Preferably, in S3, the multi-objective collaborative optimization model is realized by a weighted comprehensive objective function, which integrates the total energy consumption function of the vehicle team and the load balance function, and incorporates the real-time calculated resonance risk coefficient as a penalty term;

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

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

[0019] extracting the main frequency and harmonic components of the real-time vibration 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.

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

[0021] When it is determined that the target speed set value of a certain 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, while keeping the total operation target unchanged.

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

[0023] When the vibration frequency coupling of multiple fracturing trucks is detected 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 frequency of the remaining vehicles by more than a safe frequency difference; and the speed set value of the adjustment objects is recalculated.

[0024] Preferably, during the execution 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 an instruction to forcibly reduce the speed to a safe speed for the specific fracturing truck; the central cooperative controller immediately executes S8 to replace any final control instruction to be sent with the safety instruction and directly sends it to the corresponding local controller.

[0025] Preferably, the fault tolerance strategy further includes an execution process of:

[0026] During the execution of S1 to S7, the equipment health state code and performance parameters of each fracturing truck are monitored in real time; when it is monitored that a fracturing truck has a fault or a continuous abnormality in a key performance parameter, it is immediately excluded from the current cooperative vehicle fleet model, and based on the state of the remaining vehicles, the execution is restarted from S2 to generate new final control instructions.

[0027] Preferably, the total target includes a total target displacement and a total target pressure; the total target displacement is calibrated by a wellhead flowmeter, and the total target pressure is determined according to a wellhead pressure sensor and a formation breakdown pressure curve.

[0028] Preferably, the generation process of the equipment health state code includes:

[0029] Engine cylinder pressure, lubricating oil pressure, transmission oil temperature, and bearing vibration energy parameters are collected; each parameter is compared with a preset threshold value, and a comprehensive state score is generated according to the items and degrees that exceed the threshold value; and the score is mapped to a predefined equipment health state code.

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

[0031] The current state is taken as an initial condition to predict the state change of the vehicle fleet in a future finite time domain; the optimization problem in the time domain is solved to obtain an optimal control sequence; and the first control amount in the sequence, i.e., the target speed set value, is output for execution.

[0032] After the above technical solutions are adopted, the application has the following beneficial effects:

[0033] The fracturing fleet power collaborative distribution and risk avoidance method of the application is executed by a central collaborative controller, which is in communication connection with local controllers corresponding to the multiple fracturing vehicles; the method comprises the following steps: collecting the running state parameters of each fracturing vehicle in real time and receiving the total target of the operation; the running state parameters include engine speed, output torque, fracturing pump outlet pressure, fracturing pump outlet flow, vibration spectrum characteristics and device health status code; based on the collected running state parameters, a dynamic collaborative state model of the fleet is constructed, which is used to represent the real-time capability boundary of each fracturing vehicle and the interaction between vehicles; a multi-objective collaborative optimization model is established with the total target of the operation as the primary constraint condition, and the minimum total energy consumption of the fleet, the maximum load balancing degree of the fleet and the minimum system resonance risk coefficient as the optimization objectives; the multi-objective collaborative optimization model is solved in real time to calculate the target speed set value of each fracturing vehicle in the next control cycle; it is judged whether any target speed set value calculated is close to the preset safe working limit of the corresponding fracturing vehicle; based on the vibration spectrum characteristics of each fracturing vehicle, it is judged whether there is a potential risk of inducing system resonance in the fleet; according to the judgment result, the target speed set value is adjusted by selecting the corresponding risk avoidance strategy, the risk avoidance strategy includes load redistribution strategy and active frequency offset strategy, and the final control instruction is generated based on the adjusted result; the final control instruction is distributed to the local controllers of the fracturing vehicles, and the local controllers control the corresponding fracturing vehicles to execute the final control instruction.

[0034] In the process of collaborative distribution and risk avoidance, first, the leap from single machine independent control to fleet system optimization is realized, through global optimization and dynamic load balancing, the overall fuel consumption or electric energy consumption of the fleet can be significantly reduced under the condition of meeting the same construction requirements, and the operation economy is improved. Secondly, through real-time vibration monitoring and resonance risk modeling, the multi-machine resonance risk that cannot be detected by traditional methods can be warned and actively avoided, the operation safety is enhanced, and the expensive equipment is protected. Thirdly, when a single vehicle fails or performance drops, the system can automatically and quickly redistribute the task, realize seamless degradation operation, and ensure the continuity and reliability of the fracturing operation. In summary, the application not only improves the efficiency and economic benefit of fracturing operation, but more importantly, an intelligent risk warning and avoidance system is built at the system level, which provides core technical support for realizing safe, efficient and self-adaptive modern fracturing construction. BRIEF DESCRIPTION OF DRAWINGS

[0035] The application will be further described below in combination with the drawings and examples.

[0036] Figure 1 is a flowchart of the fracturing fleet power collaborative distribution and risk avoidance method of the embodiment of the application. DETAILED DESCRIPTION

[0037] In order to make the objects, technical solutions and advantages of the present application clearer, the present application will be further described in detail below with reference to the drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and not used to limit the present application.

[0038] As shown in Figure 1 The fracturing fleet power collaborative allocation and risk avoidance method of the present application is executed by a central collaborative controller, which is in communication connection with local controllers corresponding to each fracturing truck; the method comprises the following steps:

[0039] S1, real-time collection of running state parameters of each fracturing truck and reception of total job target; the running state parameters include engine speed, output torque, fracturing pump outlet pressure, fracturing pump outlet flow, vibration spectrum characteristics and device health status code;

[0040] S2, construction of a dynamic collaborative state model of the 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 interactive influences between vehicles;

[0041] S3, establishment of a multi-objective collaborative optimization model with achievement of the total job target as the primary constraint condition, the multi-objective collaborative optimization model taking minimization of total energy consumption of the fleet, maximization of load balance degree of the fleet and minimization of system resonance risk coefficient as optimization objectives;

[0042] S4, real-time solution of the multi-objective collaborative optimization model, calculation of target speed set values of each fracturing truck in the next control cycle;

[0043] S5, judgment of whether any target speed set value calculated is close to a preset safe working limit of the corresponding fracturing truck;

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

[0045] S7, selection of 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 generation of a final control instruction based on the adjusted result;

[0046] S8, distribution of the final control instruction to local controllers of each fracturing truck, control of the corresponding fracturing truck by each local controller to execute the final control instruction.

[0047] In this application, the central collaborative controller does not simply command each fracturing truck individually, but rather views the entire fleet as a dynamically controllable organic whole. First, two types of key information are obtained in real time through S1: one is the all-around running state parameters including mechanical vibration spectrum and equipment health status code; the other is the total displacement and total pressure that need to be reached. In S2, the central collaborative controller uses these real-time running state parameters to build a dynamic collaborative state model of the fleet. This model not only quantifies the upper limit of the ability of each truck at the current time (such as the maximum safe speed based on the health status), but also estimates the mutual mechanical influence between vehicles through shared pipelines and infrastructure, laying a precise numerical foundation for global decision-making.

[0048] Based on the dynamic collaborative state model, a multi-objective optimization problem is established in S3. The primary task is to meet the total displacement and total pressure required for construction under this hard constraint, while pursuing three optimization objectives: reducing the fuel or energy consumption of the entire fleet, balancing the load between vehicles, and minimizing the risk of system resonance caused by multi-machine vibration coupling. S4 solves the above complex optimization problem quickly and in real time, and finally outputs a set of individual target speed setting values for each fracturing truck. S5 and S6 perform double safety checks on this preliminary scheme: one is to check whether any vehicle is assigned a speed task that exceeds its own safety working limit; the other is to determine from the vibration spectrum characteristics whether there is a potential risk of triggering overall structure resonance if the fleet operates at this speed.

[0049] After safety screening, in S7, the central collaborative controller intelligently selects and triggers the corresponding risk avoidance strategy. If a vehicle has too high a load, the load redistribution strategy is started, which automatically transfers part of its load to other vehicles with sufficient capacity while ensuring the total target remains unchanged; if a resonance risk is detected, the active frequency offset strategy is started, which fine-tunes the speed of related vehicles to avoid dangerous frequency bands. After these fine adjustments, the final safe and optimized collaborative control instructions are generated. Finally, in S8, these instructions are distributed to the local controllers of each vehicle for execution, thereby converting the central optimization decision into actual actions of the fleet. The entire process starts again in the next control cycle (usually seconds), forming a continuous closed loop of perception-modeling-optimization-checking-adjustment-execution, achieving real-time tracking and dynamic optimal control of the fleet state.

[0050] Firstly, the application realizes the paradigm shift from single-machine independent control to fleet system optimization. Through global optimization and dynamic load balancing, the overall fuel consumption or power consumption of the fleet can be significantly reduced under the condition of meeting the same construction requirements, improving the economic efficiency of the operation. Secondly, through real-time vibration monitoring and resonance risk modeling, the application can warn and actively avoid the multi-machine resonance risk that cannot be detected by traditional methods, greatly enhancing the safety of the operation and protecting expensive equipment. Thirdly, the fault tolerance and emergency mechanism of the application greatly improves the robustness of the system; when a single vehicle fails or performance drops, the system can automatically and quickly redistribute tasks, enabling seamless degraded operation, ensuring the continuity and reliability of the fracturing operation. In summary, the application not only improves the efficiency and economic benefits of fracturing operations, but more importantly, it builds an intelligent risk warning and avoidance system at the system level, providing core technical support for safe, efficient, and self-adaptive modern fracturing construction.

[0051] In the application, the real-time collection of the operating state parameters of each fracturing vehicle in S1 and the reception of the total target of the operation are the starting point of data perception and input for the closed-loop cooperative control realized by the application. The core lies in the synchronous acquisition of two types of information, which provides a basis for all subsequent analysis, decision-making, and control.

[0052] The first type of information is the refined operating state parameters of each fracturing vehicle, which is a multi-dimensional data set, specifically including: engine speed and output torque to represent the instantaneous output capacity and working condition of the power system, which is the direct basis for calculating power and energy consumption; fracturing pump outlet pressure and fracturing pump outlet flow directly reflect the execution effect of the fracturing operation, which are the core indicators for evaluating whether the process requirements are met; vibration frequency spectrum characteristics are obtained by frequency domain analysis of mechanical vibration signals, which can reveal the operating frequency, harmonic component, and energy distribution of rotating parts, and are the key dynamic characteristics for evaluating mechanical state and predicting system resonance risk; the equipment health state code is a comprehensive evaluation result after preprocessing, which quantifies the overall health and performance degradation of the equipment in the form of code based on the analysis of multi-source sensor data, and directly defines the reliable working boundary.

[0053] The second type of information is the total target of the operation, which comes from the upper layer design and indicates the final flow and pressure requirements that need to be achieved by the entire fleet cooperative operation, which is the total constraint and guide for all optimization and control.

[0054] The collection of the operating state parameters is completed by the local controllers installed on each fracturing truck and the sensor network connected thereto, wherein the engine speed and the output torque are directly read from the engine electronic control unit through the controller area network bus, the fracturing pump outlet pressure and the fracturing pump outlet flow are measured by the pressure transducer and the flow meter installed on the high-pressure manifold and the signals are transmitted to the local controller, the vibration spectrum features are collected by the vibration acceleration sensor, and then the signals are converted by the data acquisition card and the real-time spectrum features are extracted by the local controller through the fast Fourier transform algorithm, and the equipment health state code is not the original signal, but a comprehensive evaluation code generated online by the local controller by comprehensively analyzing the engine cylinder pressure, lubricating oil temperature and other multi-source monitoring parameters and combining the threshold rule; all the collected or generated parameters are uploaded to the central collaborative controller by the local controllers through the industrial Ethernet or wireless private network according to the preset protocol.

[0055] At the same time, the receiving of the total operation target is the process of interaction between the central collaborative controller and the upper system, which is mainly realized through two ways: one is that the operator manually inputs or confirms the total displacement and total pressure target value on the man-machine interface in the control room according to the construction design, and the other is that the central collaborative controller automatically receives the instructions from the superior fracturing construction optimization and command system according to the real-time geological engineering model. The total operation target includes the total target displacement and the total target pressure; the total target displacement is calibrated by the wellhead flow meter, and the total target pressure is determined according to the wellhead pressure sensor and the formation breakdown pressure curve.

[0056] In summary, the collection of the operating state parameters is a bottom-up distributed sensing process, and the receiving of the total operation target is a top-down instruction issuing process, both of which converge to the central collaborative controller through a reliable industrial communication network, and together constitute the information input basis of the whole collaborative optimization and control method.

[0057] In S2, the dynamic collaborative state model of the truck fleet is constructed, that is, a digital model for describing the dynamic characteristics of the truck fleet system is established by using the collected operating state parameters. The model calculates and updates two core elements in real time through data processing and system identification: the real-time capability boundary of each fracturing truck and the interaction between vehicles.

[0058] Specifically, in S2, the construction of the dynamic collaborative state model includes the following steps:

[0059] S21, a single-machine dynamic capability boundary model based on the health state is constructed;

[0060] For each fracturing truck, the maximum power range allowed to be output by the truck in the current control period is determined according to the equipment health state code and the real-time working condition; including:

[0061] The central cooperative controller reads the device health status code uploaded by each local controller The first fracking truck) and obtains the corresponding derating coefficient according to a preset health status code and derating coefficient mapping relationship table

[0062] In combination with the universal characteristic curve data of the engine of the fracking truck (including the rated rotating speed and the rated torque ), and the real-time collected transmission oil temperature and lubricating oil pressure, the current allowed maximum rotating speed and maximum torque of the truck are calculated; the formulae are as follows:

[0063] =

[0064] =

[0065] Among them, and are working condition correction factors based on real-time oil temperature and oil pressure;

[0066] The current allowed maximum output power is determined:

[0067] =

[0068] Among them, is a comprehensive performance correction coefficient based on the current working condition, used for dynamically derating the theoretical maximum power;

[0069] The allowed working domain set of each fracking truck is output as the real-time capability boundary in the dynamic cooperative state model;

[0070] S22, construct a water force interaction influence matrix of the truck fleet;

[0071] Define a water force sensitivity coefficient , representing the influence degree of the flow change of the first fracking truck on the pump port pressure of the first fracking truck, and the formula is as follows:

[0072]

[0073] Through real-time collected outlet flow and system pressure data of each truck, a water force interaction matrix is constructed, reflecting the above 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; The capability boundary constraint set of each vehicle; the model is used as a prediction model for subsequent model predictive control (MPC). Specifically, in S4, a model predictive control algorithm is used to solve the multi-objective collaborative optimization model in real time. The steps of the algorithm are to use the prediction model (i.e., the dynamic collaborative state model constructed in S2) to predict the future state and solve the optimization with the current state as the initial condition, and to obtain the target speed set value of each fracturing vehicle in the next control period.

[0083] The determination of the real-time capability boundary depends on the dynamically input device health state code and the current working condition. For example, a health state code indicating performance degradation will cause the model to lower the allowed maximum torque and speed threshold of the engine of the vehicle. At the same time, combined with real-time parameters such as the current engine speed and output torque, the model can evaluate the thermal load and mechanical load, and thus dynamically calculate the upper limit of the power and speed that the vehicle can safely output in the next period. This makes the capability boundary a dynamic value that changes with the state.

[0084] The interaction between vehicles mainly includes mechanical vibration coupling and hydraulic system coupling. For mechanical coupling, the model analyzes the proximity and energy of the main frequency based on the vibration spectrum characteristics of each fracturing vehicle, and calculates the transmission and superposition effect of vibration energy in the vehicle fleet structure combined with the physical layout of the vehicle. For hydraulic coupling, the model estimates the back pressure disturbance caused to other vehicles when the output of a fracturing vehicle changes based on the pump outlet pressure, flow rate of each fracturing vehicle, and shared manifold network topology, and uses fluid mechanics principles. Finally, the model outputs a system state description including the capability boundaries of all vehicles and the interaction matrix, providing accurate system-level constraints for subsequent global optimization, so that the central controller can perform overall collaborative control of the vehicle fleet.

[0085] In S3, the multi-objective collaborative optimization model is realized by a weighted comprehensive objective function, which integrates the total energy consumption function of the vehicle fleet and the load balancing function based on the real-time calculated resonance risk coefficient as a penalty term; the total energy consumption function of the vehicle fleet is calculated according to the engine universal characteristic curve and the current working condition of each fracturing vehicle; the load balancing function is obtained by evaluating the deviation of the actual load of each vehicle from the average load; and 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 realized by a weighted comprehensive objective function. Firstly, two objective functions that need to be minimized are integrated: the total energy consumption function of the fleet and the load balance degree function of the fleet; secondly, the resonance risk coefficient calculated in real time is taken as a penalty term. The calculation of the total energy consumption function of the fleet is based on the universal characteristic curve of the engine of each fracturing truck and the real-time working condition to quantify the fuel consumption under different speeds and torques. The load balance degree function is obtained by evaluating the deviation between the actual load of each truck and the average load of the fleet, and the purpose is to make the load of each truck as uniform as possible. The resonance risk coefficient is calculated based on the frequency domain information analyzed in real time from the vibration frequency spectrum characteristics, which quantifies the possibility and severity of resonance caused by the coupling of the vibration frequencies of multiple vehicles. Through this design, the weighted comprehensive objective function can guide the optimization solving process to find a vehicle fleet control strategy that achieves the best trade-off among reducing total fuel consumption, balancing the load of each vehicle, and avoiding the risk of system resonance while ensuring that the total target of the job is met.

[0087] The calculation steps of the resonance risk coefficient include:

[0088] Extract the main frequency and harmonic components of the real-time vibration frequency spectrum of each fracturing truck; calculate the frequency difference f between the vibration main frequencies of any two fracturing trucks; when the frequency difference f is less than the 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.

[0089] In S7, the execution process of the load redistribution strategy includes: when it is determined that the target speed set value of a certain 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, keeping the total target of the job unchanged.

[0090] When the central collaborative controller determines that the target speed set value of a certain fracturing truck exceeds the preset proportion of its safe working limit, the truck will be marked as a load-limited object. Subsequently, the central collaborative controller recalculates the target speed set values of the remaining fracturing trucks in the fleet based on the original multi-objective collaborative optimization model. The goal of the recalculation is to make these vehicles share the load increment originally borne by the load-limited object within their respective safe working ranges, thereby keeping the total displacement and total pressure targets of the job unchanged as a whole. This process dynamically adjusts the output of each vehicle through optimization algorithms, realizes the safe transfer of load from overloaded vehicles to vehicles with surplus capacity, and ensures the continuity and safety of collaborative operation of the vehicle fleet.

[0091] In S7, the execution process of the active mistuning strategy includes: when it is detected that the vibration dominant 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 running frequency of the adjustment objects after adjustment is separated from the dominant frequencies of the remaining vehicles by a safe frequency difference; and the speed set value of the adjustment objects is recalculated.

[0092] When it is detected that the vibration dominant frequencies of multiple fracturing trucks are coupled and the calculated resonance risk coefficient exceeds the preset threshold value, the central collaborative controller selects one or more of the vehicles as adjustment objects; then, a specific frequency offset is set for the selected adjustment objects, so that the running frequency of the adjustment objects after adjustment is separated from the dominant frequencies of the remaining vehicles in the vehicle fleet by a safe frequency difference; finally, the target speed set value of the adjustment objects is recalculated according to the frequency offset. By actively and purposefully adjusting the running frequency of the potential resonance source vehicle, the conditions for frequency coupling of multiple machines are fundamentally destroyed, thereby preventing the occurrence of systemic resonance at the vibration mechanics level; the system dynamic stability in the high-power-density, multiple-machine parallel operation scenario is improved, structural fatigue damage, seal failure, and even equipment chain failure caused by resonance are avoided, the safety and reliability of fracturing operations are ensured, and through accurate recalculation of the speed, the core operation indicators such as construction displacement and pressure are not affected.

[0093] It is worth noting that identifying frequency coupling alone is not enough to determine the level of resonance risk, and quantitative evaluation is also required. The main basis is: first, the closeness of the coupling, the smaller the frequency difference f, the tighter the coupling, and the higher the risk base value; second, the size of the vibration energy involved in the coupling, the larger the amplitude corresponding to the vibration dominant frequency, the greater the excitation energy it contains, and the more intense the resonance response it may cause. The frequency difference and vibration energy are integrated through a pre-defined model or algorithm to calculate the specific system resonance risk coefficient under the current state. Finally, the system resonance risk coefficient is compared with the preset safe threshold and the higher emergency threshold to determine the risk level, which includes no risk, potential risk (which needs to be monitored or adjusted), or emergency risk (which needs to be immediately interrupted).

[0094] The application also includes an emergency protection mechanism. During the execution 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, the subsequent steps are immediately interrupted, and an independent safety instruction is generated; the safety instruction includes an emergency shutdown instruction for a specific fracturing truck or an instruction to forcibly reduce the speed to a safe speed; 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.

[0095] When a single fracturing truck is identified to be about to exceed the hard safety limit in the cooperative optimization process, or the system is facing the risk of resonance of transient surge, the emergency protection mechanism forcibly jumps out of all normal optimization calculation and strategy adjustment cycles, thereby completely avoiding the protection delay that may be caused by continuous calculation or strategy iteration in an emergency. By generating a completely independent and clear safety instruction (such as emergency shutdown or forced speed reduction), and directly executing the instruction distribution step by the central cooperative controller. It is ensured that the control instruction flow is simple, unique and clear in direction at the critical moment, and the rule of replacing any final control instruction to be sent with a safety instruction eliminates the possibility of any confusion or competition between the safety instruction and the optimization instruction that may have existed risks. A clear safety hierarchy is formed with load optimization adjustment and active frequency avoidance: the former is performance optimization within the safety boundary, and the emergency protection mechanism is the ultimate protection measure when the safety boundary is about to be broken or a systemic crisis has been detected. Therefore, the emergency protection mechanism as the safety bottom line guarantee of the entire cooperative control system enhances the survivability and reliability of the system under extreme working conditions or sudden abnormalities, and ensures the closed loop of risk management.

[0096] The application also includes a fault tolerance strategy; the execution process of the fault tolerance strategy includes: monitoring the equipment health status code and performance parameters of each fracturing truck in real time during the execution process of S1 to S7; when it is monitored that the fracturing truck has a fault or a continuous abnormality of a key performance parameter, it is immediately excluded from the current cooperative truck model, and based on the state of the remaining vehicles, the execution is restarted from S2 to generate new final control instructions.

[0097] The fault tolerance strategy of the application, in the continuous running process of S1 to S7, performs an independent real-time monitoring task in parallel, and the monitoring object is the equipment health status code and key performance parameters of each fracturing truck.

[0098] When the monitoring logic determines that a fracturing truck has a fault or a continuous abnormality of a key performance parameter, indicating that the truck cannot reliably perform the established cooperative task, the strategy is immediately started. The core operation is to exclude the faulty vehicle from the dynamic cooperative state model representing the overall state of the truck team, which means that the subsequent optimization calculation will no longer consider any ability contribution or interaction of the vehicle. Subsequently, the system takes the remaining vehicles with normal state as the object, and restarts the complete cooperative optimization and control instruction generation process from S2.

[0099] The fault-tolerant strategy endows the whole cooperative control system with the ability of online self-adaptive reconfiguration and degraded operation. Firstly, the fault-tolerant strategy realizes the active isolation and system reconfiguration of faults, which prevents the abnormal state of the faulty unit from polluting the global optimization results or sending incorrect instructions to other vehicles, and guarantees the cleanliness of the core control logic. Secondly, the fault-tolerant strategy ensures the continuity of the operation, and in the case of partial equipment failure, the system automatically and quickly recalculates a feasible and optimized cooperative scheme based on the remaining resources, so that the vehicle fleet can continue to maintain the operation under reduced total capacity, thereby minimizing unplanned downtime. Finally, the fault-tolerant strategy improves the overall availability and robustness of the system, enabling it to handle unexpected equipment conditions in actual operations with ease, localizing the impact of single-point failures and achieving smooth degradation of system functions. Therefore, the fault-tolerant strategy is not a separate remedial measure, but a key resilience design that is integrated into the main cooperative optimization process and ensures the long-term reliable operation of the system.

[0100] The generation process of the equipment health state code includes: collecting engine cylinder pressure, lubricating oil pressure, gearbox oil temperature, bearing vibration energy parameters; comparing each parameter with the preset threshold, generating a comprehensive state score according to the items and degree exceeding the threshold; mapping the score to a predefined equipment health state code.

[0101] Firstly, the key parameters such as engine cylinder pressure, lubricating oil pressure, gearbox oil temperature and bearing vibration energy are collected. Then, the measured value of each parameter is compared with the preset threshold. The system calculates a comprehensive state score according to the parameter items that exceed the limit, the amplitude of exceeding the limit and the duration, and according to the established rules. Finally, the score is converted into a standardized equipment health state code (for example, using numbers 0 to 3 to represent healthy, warning, performance degradation and serious failure) through a preset mapping relationship. This state code provides a clear quantitative index of equipment health for the central cooperative controller, which is directly used to dynamically define the safety working capacity boundary of the vehicle, 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: predicting the state change of the vehicle fleet in a finite time domain in the future with the current state as the initial condition; solving the optimization problem in this time domain to obtain the optimal control sequence; and outputting the first control quantity in the sequence, i.e. the target speed set value, for execution.

[0103] The model predictive control algorithm operates based on the principle of rolling optimization and feedback correction. Firstly, the current dynamic cooperative state of the vehicle fleet is taken as the initial condition to predict the system state variation in a finite time domain in the future. Then, a multi-objective optimization problem is solved in the prediction time domain to obtain an optimal control sequence. Finally, only the first control amount in the sequence, i.e. the target speed setting value, is output for execution. After each control cycle, the system updates the initial condition according to the latest measured state and repeats the above process.

[0104] The optimization process can consider the subsequent effects of control actions, thereby achieving smoother and more reasonable cooperative decisions while meeting 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 and stable execution of the cooperative control strategy in complex dynamic environments.

[0105] The above only describes the preferred embodiments of the present application and is not intended to limit the present application. Any modification, equivalent replacement and improvement made within the spirit and principle of the present application shall be included in the protection scope of the present application.

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 comprise 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 trucks; S3, establishing a multi-objective collaborative optimization model with the total job target as a 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 a system resonance risk coefficient as optimization objectives; S4, solving the multi-objective collaborative optimization model in real time to calculate target speed set values of each fracturing truck in a next control period; S5, judging whether any target speed set value calculated is close to a preset safe working limit of the corresponding fracturing truck; S6, judging whether the truck fleet has a potential risk of inducing system resonance based on vibration frequency spectrum characteristics of the fracturing trucks; S7, selecting a corresponding risk avoidance strategy according to the judgment results of S5 and S6 to adjust the target speed set values, the risk avoidance strategy comprising a load redistribution strategy and an active frequency offset strategy, and generating a final control instruction based on an adjusted result; S8, distributing the final control instruction to 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, In S3, the multi-objective collaborative optimization model is realized by a weighted comprehensive objective function, which integrates a total energy consumption function and a load balance degree function of the truck fleet, and incorporates a real-time calculated resonance risk coefficient as a penalty term; wherein the total energy consumption function is calculated according to engine universal characteristic curves and current working conditions of the fracturing trucks; the load balance degree function is obtained by evaluating deviations of actual loads of the trucks from an average load; and the resonance risk coefficient is calculated in real time based on frequency domain analysis of the vibration frequency spectrum.

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

4. The frac fleet power co-assignment and risk aversion method of claim 1, wherein, In S7, the execution process of the load redistribution strategy comprises: when it is determined that a target speed set value of a fracturing truck exceeds a preset proportion of a safe working limit thereof, marking the fracturing truck as a load-limited object; based on the multi-objective collaborative optimization model, recalculating target speed set values of other fracturing trucks to make them share load increments of the load-limited object within a safe range, and keeping the total job target unchanged.

5. The frac fleet power co-assignment and risk aversion method of claim 1, wherein, In S7, the execution process of the active frequency offset strategy comprises: When the coupling of the vibration main frequencies of multiple fracturing trucks is detected 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 a frequency difference greater than the safe frequency difference; and the speed set value of the adjustment objects is recalculated.

6. The frac fleet power co-assignment and risk aversion method of claim 1, wherein, During the execution 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 an instruction to forcibly reduce the speed to a safe speed for the specific fracturing truck; the central cooperative controller immediately executes S8 to replace any final control instruction to be sent with the safety instruction and directly sends it to the corresponding local controller.

7. The frac fleet power co-assignment and risk aversion method of claim 1, wherein, A fault tolerance strategy is also included. The execution process of the fault tolerance strategy includes: During the execution of S1 to S7, the equipment 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 that a key performance parameter is continuously abnormal, the fracturing truck is immediately removed from the current cooperative truck model, and based on the state of the remaining vehicles, the execution is restarted from S2 to generate new final control instructions.

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

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

10. The method of frac fleet power co-assignment and risk aversion of claim 1, wherein, 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: The current state is taken as the initial condition to predict the state change of the truck fleet in a future finite time domain; the optimization problem in this time domain is solved to obtain an optimal control sequence; The first control amount in the sequence, i.e., the target speed set value, is output for execution.

Images