Generalized inertial precision measurement and decoupling system and method for complex engineering equipment

Abstract

The present application belongs to the field of generalized inertia measurement and decoupling system of complex engineering equipment, and discloses a generalized inertia precise measurement and decoupling system and method of complex engineering equipment. The present application proposes an energy-based inertia measurement decoupling framework. According to the form of energy transmission in construction equipment, the generalized inertia support energy physical expression method of electrical, hydraulic and mechanical systems is determined. The inertia contribution of different systems is decomposed by using generalized inertia relaxation factors. In order to realize accurate prediction of generalized inertia support energy, an encoder-decoder neural network integrated with compression excitation and shared weight strategy is introduced. The inertia decoupling analysis model is established by combining the physical driving model with the data-driven model, and the generalized inertia relaxation factor is quantified. The reliability of the method is verified by experiments.

Description

Technical Field

[0001] The present invention belongs to the field of generalized inertia measurement and decoupling systems for complex engineering equipment, and in particular relates to a generalized inertia precision measurement and decoupling system and method for complex engineering equipment. Background Art

[0002] With the advancement of urbanization and the development of the construction industry, construction machinery is increasingly moving towards high efficiency, energy conservation, and environmental protection. Construction equipment, particularly intelligent equipment such as tunnel boring machines (TBMs) and aerial building machines, is indispensable in various large-scale construction projects due to its high integration and efficient operation. Currently, construction equipment is often subject to the complex coupling of electrical, hydraulic, and mechanical systems and a large number of uncertain disturbances during operation. "Time lag" or "cyclic vibration" is common in construction equipment operation, especially under the changing working environment and complex operating parameters. Due to the inherent inertia of electrical, hydraulic, and mechanical systems, the relevant control schemes often fail to respond in real time, which can easily lead to equipment damage and construction accidents. In 2020, untimely response of lifting equipment at a construction site in Shanxi Province, China resulted in the deaths of three construction workers and direct economic losses of over 4 million yuan. In 2022, during the construction of a Yokohama subway tunnel in Japan, bolts on a tunnel boring machine broke due to inherent inertia and cyclic disturbances, delaying construction by seven months. Therefore, it is necessary to explore in detail the identification, measurement and control of the inertia effect to ensure the efficient, safe and reliable operation of construction equipment.

[0003] Inertia is the inherent property of a system that resists changes in state caused by external disturbances. First proposed by Galileo and summarized in Newtonian mechanics, inertia refers to the property of an object maintaining its state of motion, representing the difficulty of changing its state. The presence of inertia prevents external responses from being reflected in the system in real time. Specifically, inertia is the property of a particular type of object, system, or organization to maintain a specific state. In addition to mechanical systems, inertia has also been extensively studied in thermodynamics, management, and power systems. Thermal inertia describes how quickly surface temperature fluctuations decay when one side of an object is subjected to periodic thermal work. Management inertia refers to the efficiency with which an enterprise, government, or collective responds to directives. Inertia in power systems refers to the ability of various energy sources to resist frequency changes caused by active disturbances. Different application scenarios have different requirements for inertia. For example, in a device's braking system, inertia is desired to be as low as possible, while in the thermal performance of a building, inertia is desired to be as high as possible. In summary, the beneficial or detrimental effects of inertia are determined by the application scenario, making precise measurement of inertia essential.

[0004] Construction equipment generally consists of mechanical, hydraulic, and electrical systems. Due to the inertial coupling and inertial inheritance phenomena between these different systems, the system's inertia is difficult to quantify. Furthermore, considering the interactions and feedback between each system and its environment, quantifying the impact of inertia and achieving precise control of construction equipment presents a significant challenge. To address this issue, this paper introduces the concept of generalized inertial support energy, which is the energy required to support the equipment's current operating state. Specifically, system inertia is essentially the process by which the system's existing energy is not effectively released, and the process of inertia disappearing is also the process by which generalized inertial support energy is consumed. Energy, as a measure of the temporal and spatial distribution of mass, often permeates each subsystem and even the entire operating cycle of construction equipment. Therefore, constructing a generalized inertial theoretical expression using generalized inertial support energy can further achieve precise management and control of the inertial effects of construction equipment, which is of great research value.

[0005] Through the above analysis, the problems and defects of the existing technology are as follows:

[0006] Construction equipment typically consists of mechanical, hydraulic, and electrical systems. Due to the existence of inertial coupling and inertial inheritance between these systems, system inertia is difficult to quantify. Furthermore, considering the interactions and feedback between these systems and their environment, quantifying the impact of inertia and achieving precise control of construction equipment is a major challenge. Summary of the Invention

[0007] In response to the problems existing in the prior art, the present invention provides a system and method for precise measurement and decoupling of generalized inertia of complex engineering equipment.

[0008] The present invention is implemented as follows: a system and method for precise measurement and decoupling of generalized inertia of complex engineering equipment includes:

[0009] Step 1: Decomposition and summary of electrical, hydraulic and mechanical systems in building equipment;

[0010] Analyze the basic structure of different construction equipment, determine the energy transfer paths between different systems, identify the factors affecting the inertia of the equipment and the main sources, and identify the common inertia forms in complex engineering equipment;

[0011] Step 2: Theoretical analysis and derivation of generalized inertia;

[0012] Determine the generalized inertia expressions for the electrical, hydraulic, and mechanical parts, derive formulas to construct a calculation method for the generalized inertia support energy of the entire machine, introduce a generalized inertia relaxation factor, and integrate the inertia of each subsystem to form a theoretical expression of the inertia of the entire machine;

[0013] Step 3: Measurement of generalized inertia;

[0014] A data-driven proxy model is constructed for the parameter of the generalized inertia support energy of the entire device. A deep neural network framework for encoding and decoding is designed to accurately measure the device inertia and achieve data-based generalized inertia characterization.

[0015] Step 4: Decoupling analysis of generalized inertia;

[0016] A parameter decoupling model is constructed by combining the theoretical analysis model of inertia and the data-driven model. The optimization target is designed to make the measurement results of the two consistent, thus realizing the decoupling of the inertia relaxation factors of the electrical, hydraulic and mechanical parts, and ensuring the accurate theoretical expression of the inertia of the whole machine.

[0017] Furthermore, the electrical system:

[0018] The generalized inertia support energy of the electrical system consists of two parts, namely the electric drive part and the electronic control part. The electric drive part is mainly composed of various motors. For a running motor, the inertia is expressed by the degree of resistance to speed changes. The rotational inertia of the motor is expressed as:

[0019] J = ∫r 2 dm

[0020] Where r is the radius of rotation; m is the mass of the rigid body; the unit of moment of inertia J is kgm 2 , for the motor, the moment of inertia is constant;

[0021] When an unbalanced disturbance occurs in the motor, the inertia is manifested as fluctuations in the motor output energy. Therefore, the generalized inertia support energy of the electric drive part is expressed as:

[0022]

[0023] Where J is the moment of inertia, ω1 is the mechanical angular velocity before the system disturbance, ω2 is the mechanical angular velocity after the system disturbance, T is the output torque, and t is time. The generalized inertia support energy of the motor is expressed as the energy change before and after the system disturbance, that is, the energy of the system to overcome the inertia and do work.

[0024] In the electronic control system, the generalized inertial support energy is expressed by the response changes to different electrical signals, and its expression is:

[0025] E e2 =V∫I1dt-∫I2dt

[0026] Where V is the voltage of the control circuit, I1 is the current of the control signal before the system disturbance, and I2 is the current of the control signal after the system disturbance. Compared with the electric drive system, the generalized inertia support energy of the electric control system is much smaller. In this case, the generalized inertia support energy of the electric drive system is used as the main inertia source of the electric transmission system.

[0027] Furthermore, the hydraulic system:

[0028] Due to the inertia of the driving, executing, and controlling components in the hydraulic system, as well as the fluidity and viscosity of the hydraulic oil itself, the inertia within the hydraulic system can be reflected at the output end. Therefore, the fluctuation of the hydraulic cylinder output energy is regarded as the generalized inertia support energy. The specific expression of the generalized inertia support energy in the hydraulic system is as follows:

[0029] E h =(P1-P2)·S·L

[0030] Where S is the effective area of ​​the hydraulic cylinder, P1 is the cylinder pressure before the system disturbance, P2 is the cylinder pressure after the system disturbance, and L is the effective working distance during the disturbance.

[0031] Furthermore, the mechanical system:

[0032] In mechanical systems, generalized inertia support energy primarily comes from the kinetic energy of moving parts. When the external excitation of a mechanical system changes, its generalized inertia support energy manifests as a change in kinetic energy. Since there is no electric current, hydraulic oil, or other fluid medium in a mechanical system, the quantification of its generalized inertia can be directly referenced by Newton's laws. The specific calculation formula is as follows:

[0033]

[0034] m is the mass of the moving part, I1 is the velocity of the moving part before the disturbance, and I2 is the velocity of the moving part after the disturbance; F is the thrust of the mechanical system, and x is the displacement of the mechanical system;

[0035] Through the above analysis, the present invention preliminarily determined the expression of generalized inertial support energy in electrical, hydraulic, and mechanical systems. However, due to the coupling effect of the electrical, hydraulic, and mechanical systems, the inertia of construction equipment cannot be simply regarded as the sum of the three systems. In addition, many assumptions are introduced in the process of calculating the generalized inertial support energy of electrical, hydraulic, and mechanical systems, resulting in a certain deviation between the calculated generalized inertial support energy and the actual situation. To address these issues, the generalized inertial relaxation factors λ1, λ2, and λ3 are introduced to accurately express the generalized inertial support energy of construction equipment, as shown below:

[0036]

[0037] Where, E e is the generalized inertia support energy of the electrical system, E h is the generalized inertia support energy of the hydraulic system, E mThe generalized inertia support energy of the mechanical system; since the energy transfer in electrical, hydraulic and mechanical systems is linear, there is no need to consider the inertia product phenomenon; the coupling between electrical, hydraulic and mechanical systems is reflected by the relaxation factor, where the generalized inertia support energy of one system can enhance or weaken the inertia effect of another system.

[0038] Furthermore, the measurement of generalized inertia is:

[0039] An encoder-decoder neural network model that integrates compression excitation and shared weight strategies. The model consists of an encoder and a decoder. The encoder uses convolutional layers, attention layers, compression excitation modules, and dense layers to extract high-level features of the input parameters. Both the convolutional and attention layers use a standard network structure. The compression excitation module focuses on squeezing and excitation of the extracted features. Squeezing mainly performs global average pooling on the data, while excitation reconstructs the features through full connectivity and activation functions. The calculation process of compression excitation is as follows:

[0040]

[0041] s=FC(z)

[0042] y=σ(s)⊙a

[0043] Where a is the input feature map, z is the global feature vector obtained by applying global average pooling to x, FC represents a fully connected layer, s is the learning vector that models channel dependencies, σ() is the sigmoid activation function, and ⊙ represents the dot product.

[0044] The decoding path is essentially the same as the encoding path; the decoder network uses this jointly defined feature as input; the decoder outputs the predicted generalized inert support energy; by sharing weights between the encoder and decoder, the training process is efficient and consistent; the activation function for the shared weights is ReLU; the model is trained in a supervised manner and converges by minimizing the loss function; the loss function is the mean squared error between the actual and predicted samples, as shown below:

[0045]

[0046] Where n is the number of training samples, L is the prediction error; θ is the network hyperparameter, f1(z) is the mapping function of the decoder, z is the feature vector, x i and x' ,i is the input feature;

[0047] In order to verify the goodness of the model, the output error of the prediction model is mainly measured by model evaluation indicators; commonly used evaluation indicators include the coefficient of determination (R2), root mean square error (RMSE), mean absolute error (MAE) and D2 absolute error score (D2); for RMSE and MAE, the smaller the indicator value, the higher the prediction accuracy of the model; the closer R2 and D2 are to 1, the higher the prediction accuracy of the model.

[0048] Further, the decoupling analysis of the generalized inertia:

[0049] After completing the proxy models for the generalized inertial support energy of electrical, hydraulic, and mechanical systems, the data-driven model was combined with the physical model for a decoupling analysis. Since the theoretical value of the generalized inertial support energy is basically consistent with the prediction results of the surrogate model, the decoupling problem of the generalized inertial relaxation factors λ1, λ2, and λ3 can be transformed into an optimization problem. By designing the corresponding optimization objectives and constraints, λ1, λ2, and λ3 are obtained. The optimization model is as follows:

[0050]

[0051] Variables:λ1,λ2,λ3

[0052] Constraint:

[0053] The present invention also provides a generalized inertia precision measurement and decoupling system for complex engineering equipment, comprising:

[0054] Decomposition and induction module: used to analyze the basic structure of construction equipment, determine the energy transfer path between different systems, clarify the inertia influencing factors and main sources of the electrical system, hydraulic system and mechanical system in the equipment, and the common inertia forms in complex engineering equipment.

[0055] Theoretical analysis and derivation module: used to determine the generalized inertia expressions of the electrical, hydraulic, and mechanical parts, construct a generalized inertia support energy calculation method for the entire machine, and introduce a generalized inertia relaxation factor to integrate the inertia of each subsystem to form a theoretical expression of the inertia of the entire machine.

[0056] Measurement module: Builds a data-driven proxy model for the generalized inertia support energy of the entire machine, and uses an encoding and decoding deep neural network framework to accurately measure the device inertia, thereby achieving data-based characterization of generalized inertia.

[0057] Decoupling Analysis Module: This module combines the theoretical analysis model of inertia with the data-driven model to construct a parameter decoupling model. The module designs optimization targets to ensure consistency between the two measurement results, decoupling the inertia relaxation factors of the electrical, hydraulic, and mechanical components, and ensuring accurate theoretical expression of the inertia of the entire machine.

[0058] Further, including:

[0059] Electric drive analysis module: used to analyze the motor's moment of inertia and the fluctuation of the motor's output energy when an unbalanced disturbance occurs, thereby determining the generalized inertia support energy of the electric drive part.

[0060] Electronic control part analysis module: used to analyze the response changes of the control circuit to different electrical signals to determine the generalized inertial support energy of the electronic control system.

[0061] Integration module: used to integrate the generalized inertial support energy of the electric drive part and the electronic control part, serving as the main inertial source of the electrical system.

[0062] Further, including:

[0063] Hydraulic cylinder output analysis module: used to analyze the output energy fluctuations of the hydraulic cylinder before and after system disturbances to determine the generalized inertial support energy of the hydraulic system.

[0064] Generalized inertia quantification module: used to quantify the generalized inertia support energy of the hydraulic system based on the fluctuation of the hydraulic cylinder output energy.

[0065] Further, including:

[0066] Kinetic Energy Analysis Module: It is used to analyze the changes in the kinetic energy of a mechanical system when the external excitation changes, so as to determine the generalized inertial support energy of the mechanical system.

[0067] Generalized Inertia Quantification Module: used to quantify the generalized inertia support energy of a mechanical system based on the change in kinetic energy.

[0068] In combination with the above technical solutions and the technical problems solved, the advantages and positive effects of the technical solutions to be protected by the present invention are as follows:

[0069] First, in order to analyze the inertia effect of construction equipment, the present invention proposes an energy-based inertia measurement decoupling framework. According to the form of energy transfer in construction equipment, the physical expression method of the generalized inertia support energy of electrical, hydraulic and mechanical systems is determined. The generalized inertia relaxation factor is used to decompose the inertia contribution of different systems. In order to achieve accurate prediction of the generalized inertia support energy, an encoder-decoder neural network that integrates compressed excitation and shared weight strategy is introduced. The physical drive model is combined with the data drive model to establish an inertia decoupling analysis model, and the generalized inertia relaxation factor is quantified. The reliability of the method is verified by experiments. The present invention can quantify the inertia phenomenon in traditional complex equipment, especially for dangerous forms such as untimely response and inadequate control, clarify the main parts and causes of inertia, and provide technical guarantees for precise control.

[0070] In view of the case study, the main research results are summarized as follows. (1) The constructed data-driven measurement model of generalized inert support energy has high accuracy, and the R2, RMSE, MAE and D2 of the test set are 0.987, 23.796, 14.29 and 0.957 respectively. (2) The decoupling optimization model of generalized inert support energy has good efficiency and effect, and the optimal solution of the model is 0.0018, and the optimization time is 58.4s. (3) Through the decoupling optimization model, the theoretical expression of generalized inert support energy is, which reflects the degree of inertia contribution of electrical, hydraulic and mechanical systems. (4) The theoretical expression of generalized inert support energy is verified by actual engineering samples, and the error of 81.94% of the samples is less than 5%.

[0071] Second, the present invention proposes a framework for studying the support energy of generalized inertia. By analyzing and deducing the support energy of electrical, hydraulic, and mechanical systems, a generalized inertia measurement and decoupling method is proposed. The effectiveness of the patent was verified by a tunnel boring equipment under on-site construction. The innovations of this method include: (1) a new encoder-decoder neural network was designed, which integrated compressed excitation and shared weight strategies to achieve the measurement of inertia effects. (2) a method for analyzing the inertia decoupling of electrical, hydraulic, and mechanical systems based on an optimal foraging algorithm was proposed by combining the physical drive model and the data drive model. This method breaks through engineering problems such as system decoupling and accident tracing, and provides a theoretical reference for achieving safe and reliable operation.

[0072] Third, as auxiliary evidence for the inventiveness of the claims of the present invention, it is also reflected in the following important aspects:

[0073] (1) The expected benefits and commercial value of the technical solution of the present invention after transformation are:

[0074] The technology of this invention can be transformed into an external measurement system that connects to various large-scale engineering equipment via communication protocols, providing strategic support for equipment inertia control. The relevant results can be used in collaboration with various engineering machinery manufacturers and construction companies to effectively improve system control accuracy. This field currently lacks related products and has good industrial prospects.

[0075] (2) The technical solution of the present invention fills the technical gap in the industry at home and abroad:

[0076] In response to the engineering problem of low control accuracy and weak reliability of large and complex equipment in China, a method for measuring equipment inertia is proposed. As an inherent characteristic of engineering equipment, inertia seriously affects operational safety and is particularly prone to engineering disasters caused by untimely control. For a long time, research on real-time control of engineering equipment has mainly focused on electronic and electrical systems, and people have mainly focused on how to design reliable control algorithms to ensure real-time control of equipment. From a higher perspective, the present invention comprehensively coordinates electrical, hydraulic and mechanical systems, and based on energy, an element that runs through the entire cycle of the equipment, analyzes the inert coupling between different systems, realizes the quantification of the inertia of different systems, and provides a reliable foundation for all-round control of future equipment. The relevant results can be benchmarked against the bond graph theory in Siemens' AMEsim software to construct a model of equipment control accuracy and response mechanism from the perspective of function / energy.

[0077] (3) The technical solution of the present invention solves the technical problems that people have been eager to solve but have never been able to solve successfully:

[0078] Inertia, an inherent property of equipment, has long lacked quantification. Traditional control systems have addressed this by creating margins. However, this lack of quantification often results in wasted resources or insufficient margins, posing a significant challenge to the precise control of large-scale engineering equipment. This invention decouples the inertia forms and coupling mechanisms of electrical, hydraulic, and mechanical systems from an energy perspective, achieving numerical quantification of the inertia phenomenon and providing reliable technical support for the precise design of subsequent control systems.

[0079] Fourth, the present invention relates to the performance analysis and optimization of complex engineering equipment, particularly the technology and method for precise measurement and decoupling of generalized inertia in electrical, hydraulic, and mechanical systems. This invention represents significant technological advancement in several aspects:

[0080] 1. System Decomposition and Induction: By performing a detailed decomposition and induction of the electrical, hydraulic, and mechanical systems within construction equipment, this method accurately identifies the energy transfer paths between these systems, thereby providing a deep understanding of the forms and sources of inertia within complex engineering equipment. This analytical approach not only enhances understanding of equipment performance but also provides a solid foundation for subsequent inertia measurement and decoupling.

[0081] 2. Theoretical Analysis and Derivation of Generalized Inertia: This paper constructs a method for calculating the generalized inertia support energy of the entire machine through formula derivation and introduces a generalized inertia relaxation factor. This innovative theoretical model not only integrates the inertia of each subsystem but also provides a theoretical expression of the inertia of the entire machine, providing theoretical support for subsequent measurement and decoupling.

[0082] 3. Data-Driven Agent Models and Deep Neural Networks: This paper designs a deep neural network framework for encoding and decoding, enabling data-driven, generalized, inert, and precise measurement. This data-driven approach overcomes the limitations of traditional measurement methods, improving accuracy and efficiency, and providing a powerful tool for performance evaluation and optimization of complex engineering equipment.

[0083] 4. Decoupling Analysis of Generalized Inertia: This paper combines a theoretical analysis model of inertia with a data-driven model to construct a parameter decoupling model. By designing optimization targets to ensure consistency between the measured results of the two, this method achieves decoupling of the inertia relaxation factors of the electrical, hydraulic, and mechanical components. This decoupling analysis method not only improves the theoretical accuracy of the overall inertia expression but also provides guidance for optimizing equipment performance.

[0084] This invention has achieved significant technological advancements in the precision measurement and decoupling of generalized inertia in complex engineering equipment. These advances not only enhance the understanding of equipment performance but also provide strong support for optimized equipment design and maintenance. BRIEF DESCRIPTION OF THE DRAWINGS

[0085] Figure 1 This is a flow chart of a system and method for precise measurement and decoupling of generalized inertia of complex engineering equipment provided by an embodiment of the present invention.

[0086] Figure 2 It is an overall research framework diagram of the research method provided by the embodiment of the present invention.

[0087] Figure 3 1 is a diagram of the encoder-decoder neural network structure provided by an embodiment of the present invention.

[0088] Figure 4 2 is a comparison chart of model hyperparameters provided by an embodiment of the present invention.

[0089] Figure 5 This is a comparison chart of the model loss function and optimizer provided by an embodiment of the present invention.

[0090] Figure 6 This is a diagram of the model accuracy of the training set and test set provided by an embodiment of the present invention (a) training set (b) test set.

[0091] Figure 7 Figure 1 shows the model accuracy under different construction scenarios provided by an embodiment of the present invention (a) Scene 1 (b) Scene 2 (c) Scene 3 (d) Scene 4 (e) Scene 5.

[0092] Figure 8 This is a diagram of the decoupling model optimization results provided by an embodiment of the present invention.

[0093] Figure 9This is a diagram comparing the errors between the calculated values ​​and the actual values ​​on site during construction of a tunnel boring machine according to an embodiment of the present invention. DETAILED DESCRIPTION

[0094] In order to make the purpose, technical solutions and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0095] like Figure 1 As shown, the present invention provides a system and method for precise measurement and decoupling of generalized inertia of complex engineering equipment, comprising the following steps:

[0096] S101, decomposition and summary of electrical, hydraulic and mechanical systems in building equipment;

[0097] Analyze the basic structure of different construction equipment, determine the energy transfer paths between different systems, identify the factors affecting the inertia of the equipment and the main sources, and identify the common inertia forms in complex engineering equipment;

[0098] S102, Theoretical analysis and derivation of generalized inertia;

[0099] Determine the generalized inertia expressions for the electrical, hydraulic, and mechanical parts, derive formulas to construct a calculation method for the generalized inertia support energy of the entire machine, introduce a generalized inertia relaxation factor, and integrate the inertia of each subsystem to form a theoretical expression of the inertia of the entire machine;

[0100] S103, measurement of generalized inertness;

[0101] A data-driven proxy model is constructed for the parameter of the generalized inertia support energy of the entire device. A deep neural network framework for encoding and decoding is designed to accurately measure the device inertia and achieve data-based generalized inertia characterization.

[0102] S104, Decoupling Analysis of Generalized Inertia;

[0103] A parameter decoupling model is constructed by combining the theoretical analysis model of inertia and the data-driven model. The optimization target is designed to make the measurement results of the two consistent, thus realizing the decoupling of the inertia relaxation factors of the electrical, hydraulic and mechanical parts, and ensuring the accurate theoretical expression of the inertia of the whole machine.

[0104] The present invention relates to a system and method for precise measurement and decoupling of generalized inertia in complex engineering equipment, which mainly consists of the following steps and aims to accurately measure and decouple generalized inertia in complex engineering equipment, thereby improving the performance and efficiency of the equipment.

[0105] This method combines theoretical analysis with data-driven approaches to accurately measure and interpret generalized inertia in complex engineering equipment. The theoretical analysis provides a deep understanding and mathematical representation of the factors influencing inertia, while the data-driven proxy model leverages deep learning techniques to precisely measure the inertia of actual equipment. This combination not only improves measurement accuracy and efficiency but also optimizes equipment performance through decoupled analysis, reducing unnecessary energy loss and enhancing overall engineering efficiency.

[0106] The electrical system provided by the present invention:

[0107] The generalized inertia support energy of the electrical system consists of two parts, namely the electric drive part and the electronic control part. The electric drive part is mainly composed of various motors. For a running motor, the inertia is expressed by the degree of resistance to speed changes. The rotational inertia of the motor is expressed as:

[0108] J = ∫r 2 dm

[0109] Where r is the radius of rotation; m is the mass of the rigid body; the unit of moment of inertia J is kgm 2 , for the motor, the moment of inertia is constant;

[0110] When an unbalanced disturbance occurs in the motor, the inertia is manifested as fluctuations in the motor output energy. Therefore, the generalized inertia support energy of the electric drive part is expressed as:

[0111]

[0112] Where J is the moment of inertia, ω1 is the mechanical angular velocity before the system disturbance, ω2 is the mechanical angular velocity after the system disturbance, T is the output torque, and t is time. The generalized inertia support energy of the motor is expressed as the energy change before and after the system disturbance, that is, the energy of the system to overcome the inertia and do work.

[0113] In the electronic control system, the generalized inertial support energy is expressed by the response changes to different electrical signals, and its expression is:

[0114] E e2 =V∫I1dt-∫I2dt

[0115] Where V is the voltage of the control circuit, I1 is the current of the control signal before the system disturbance, and I2 is the current of the control signal after the system disturbance. Compared with the electric drive system, the generalized inertia support energy of the electric control system is much smaller. In this case, the generalized inertia support energy of the electric drive system is used as the main inertia source of the electric transmission system.

[0116] The hydraulic system provided by the present invention:

[0117] Due to the inertia of the driving, executing, and controlling components in the hydraulic system, as well as the fluidity and viscosity of the hydraulic oil itself, the inertia within the hydraulic system can be reflected at the output end. Therefore, the fluctuation of the hydraulic cylinder output energy is regarded as the generalized inertia support energy. The specific expression of the generalized inertia support energy in the hydraulic system is as follows:

[0118] E h =(P1-P2)·S·L

[0119] Where S is the effective area of ​​the hydraulic cylinder, P1 is the cylinder pressure before the system disturbance, P2 is the cylinder pressure after the system disturbance, and L is the effective working distance during the disturbance.

[0120] The mechanical system provided by the present invention:

[0121] In mechanical systems, generalized inertia support energy primarily comes from the kinetic energy of moving parts. When the external excitation of a mechanical system changes, its generalized inertia support energy manifests as a change in kinetic energy. Since there is no electric current, hydraulic oil, or other fluid medium in a mechanical system, the quantification of its generalized inertia can be directly referenced by Newton's laws. The specific calculation formula is as follows:

[0122]

[0123] m is the mass of the moving part, I1 is the velocity of the moving part before the disturbance, and I2 is the velocity of the moving part after the disturbance; F is the thrust of the mechanical system, and x is the displacement of the mechanical system;

[0124] Through the above analysis, the present invention preliminarily determined the expression of generalized inertial support energy in electrical, hydraulic, and mechanical systems. However, due to the coupling effect of the electrical, hydraulic, and mechanical systems, the inertia of construction equipment cannot be simply regarded as the sum of the three systems. In addition, many assumptions are introduced in the process of calculating the generalized inertial support energy of electrical, hydraulic, and mechanical systems, resulting in a certain deviation between the calculated generalized inertial support energy and the actual situation. To address these issues, the generalized inertial relaxation factors λ1, λ2, and λ3 are introduced to accurately express the generalized inertial support energy of construction equipment, as shown below:

[0125]

[0126] Where, E e is the generalized inertia support energy of the electrical system, E h is the generalized inertia support energy of the hydraulic system, E mThe generalized inertia support energy of the mechanical system; since the energy transfer in electrical, hydraulic and mechanical systems is linear, there is no need to consider the inertia product phenomenon; the coupling between electrical, hydraulic and mechanical systems is reflected by the relaxation factor, where the generalized inertia support energy of one system can enhance or weaken the inertia effect of another system.

[0127] The generalized laziness measurement provided by the present invention consists of an encoder and a decoder. The encoder uses convolutional layers, attention layers, a compression-excitation module, and dense layers to extract high-level features of the input parameters. The convolutional layers and attention layers both adopt a standard network structure. The compression-excitation module focuses on squeezing and excitation of the extracted features. Squeezing mainly performs global average pooling on the data, while excitation reconstructs the features through full connectivity and activation functions. The calculation process of compression-excitation is as follows:

[0128]

[0129] s=FC(z)

[0130] y=σ(s)⊙a

[0131] Where a is the input feature map, z is the global feature vector obtained by applying global average pooling to x, FC represents a fully connected layer, s is the learning vector that models channel dependencies, σ() is the sigmoid activation function, and ⊙ represents the dot product.

[0132] The decoding path is essentially the same as the encoding path; the decoder network uses this jointly defined feature as input; the decoder outputs the predicted generalized inert support energy; by sharing weights between the encoder and decoder, the training process is efficient and consistent; the activation function for the shared weights is ReLU; the model is trained in a supervised manner and converges by minimizing the loss function; the loss function is the mean squared error between the actual and predicted samples, as shown below:

[0133]

[0134] Where n is the number of training samples, L is the prediction error; θ is the network hyperparameter, f1(z) is the mapping function of the decoder, z is the feature vector, x i and x' ,i is the input feature;

[0135] In order to verify the goodness of the model, the output error of the prediction model is mainly measured by model evaluation indicators; commonly used evaluation indicators include the coefficient of determination (R2), root mean square error (RMSE), mean absolute error (MAE) and D2 absolute error score (D2); for RMSE and MAE, the smaller the indicator value, the higher the prediction accuracy of the model; the closer R2 and D2 are to 1, the higher the prediction accuracy of the model.

[0136] Decoupling analysis of generalized inertia provided by the present invention:

[0137] After completing the proxy models for the generalized inertial support energy of electrical, hydraulic, and mechanical systems, the data-driven model was combined with the physical model for a decoupling analysis. Since the theoretical value of the generalized inertial support energy is basically consistent with the prediction results of the surrogate model, the decoupling problem of the generalized inertial relaxation factors λ1, λ2, and λ3 can be transformed into an optimization problem. By designing the corresponding optimization objectives and constraints, λ1, λ2, and λ3 are obtained. The optimization model is as follows:

[0138]

[0139] Variables:λ1,λ2,λ3

[0140] Constraint:

[0141] Reference Figure 2 As shown, this embodiment provides a technical method for real-time prediction of cylinder pressure in shield segment installation scenarios based on a trustworthy multi-head attention deep neural network guided by physical information, including the following steps:

[0142] Step 1: Decomposition and summary of electrical, hydraulic and mechanical systems in building equipment

[0143] Construction equipment typically refers to large-scale engineering equipment or systems with high energy consumption and significant carbon emissions, such as excavators and roadheaders. The primary operating mechanisms of this type of equipment typically consist of electrical, hydraulic, and mechanical systems. These three systems work together to achieve operational objectives. Specifically, the electrical system controls the operation, while the hydraulic system drives the mechanical system to complete the task. The inertia of construction equipment typically consists of three components: electrical inertia, hydraulic inertia, and mechanical inertia. The energy transfer path is the process of transferring electrical energy to hydraulic energy and then to mechanical energy.

[0144] The electrical system is the power source and main control unit of the equipment, primarily divided into two parts: the electric drive and the electronic control. The electric drive refers to the drive system of the construction equipment, primarily consisting of components such as the motor. The motor drives the hydraulic pump, converting electrical energy into hydraulic energy. The electronic control refers to the control circuitry of the entire construction equipment, utilizing various electronic components to enable start / stop, parameter adjustment, and operational protection.

[0145] In construction equipment, hydraulic systems mediate its operation. Compared to equipment where electrical systems directly control mechanical systems, hydraulic systems provide certain buffering, protection, and continuously variable speed capabilities. Because they can output high thrust or torque, hydraulic systems can achieve low-speed, high-tonnage movement, making them indispensable in construction equipment. Hydraulic systems primarily consist of drive elements (hydraulic pumps), actuators (hydraulic cylinders and hydraulic motors), a medium (hydraulic oil), and control elements (various hydraulic valves). The hydraulic pump drives the flow of hydraulic oil, enabling the operation of the cylinders and motors.

[0146] Mechanical systems are the terminals of equipment, interacting with objects to complete operations, such as excavator buckets and tunnel boring machine cutterheads. They are directly connected to the hydraulic system, which drives the corresponding operations. The mechanical systems of different equipment often vary significantly, and the equipment here is primarily distinguished by their mechanical systems. Table 1 lists several common types of construction equipment and their corresponding mechanical operating mechanisms.

[0147] Electrical, hydraulic, and mechanical systems are essential in construction equipment. Their coordinated operation enables a wide range of engineering requirements for high loads and complex working conditions. However, due to the large size and weight of construction equipment, significant inertia can occur during operation, easily leading to problems such as insufficient response and delayed braking. Therefore, there is an urgent need to quantify the inertia of each system in construction equipment for precise control.

[0148] Table 1. Typical construction equipment and corresponding actuators.

[0149]

[0150] Step 2: Theoretical analysis of generalized inertia;

[0151] The problem of generalized inertia decoupling in electrical, hydraulic, and mechanical systems is essentially a way to express inertia in different energy forms. From an energy perspective, it can be unified into the resistance of different carriers to external responses. Generalized inertia support energy, a parameter that permeates the entire operational process of the equipment, can quantify the inertia effects of each system. In this section, the generalized inertia support energy is theoretically derived from various aspects of electrical, hydraulic, and mechanical systems.

[0152] (1) Electrical system

[0153] The generalized inertia support energy of an electrical system consists of two parts: the electric drive and the electronic control. The electric drive, represented by the motor, has a higher energy density, which means more generalized inertia support energy. For a running motor, inertia is expressed as the degree of resistance to speed changes. The motor's rotational inertia is expressed as:

[0154] J = ∫r2 dm

[0155] Where r is the rotation radius; m is the mass of the rigid body; the unit of the moment of inertia J is kgm2. For the electric motor, the moment of inertia is constant.

[0156] When an unbalanced disturbance occurs in the motor, the inertia is manifested as fluctuations in the motor output energy. Therefore, the generalized inertia support energy of the electric drive part is expressed as:

[0157]

[0158] Where J is the moment of inertia, ω1 is the mechanical angular velocity before the system disturbance, ω2 is the mechanical angular velocity after the system disturbance, T is the output torque, and t is time. The generalized inertia support energy of the motor is expressed as the energy change before and after the system disturbance, that is, the energy of the system to overcome inertia and do work.

[0159] In the electronic control system, the generalized inertial support energy is expressed by the response changes to different electrical signals, and its expression is:

[0160] E e2 =V∫I1dt-∫I2dt

[0161] Where V is the voltage of the control circuit, I1 is the current of the control signal before the system disturbance, and I2 is the current of the control signal after the system disturbance. Compared with the electric drive system, the generalized inertia support energy of the electric control system is much smaller. In this case, the generalized inertia support energy of the electric drive system is used as the main inertia source of the electric transmission system.

[0162] (2) Hydraulic system

[0163] Due to the inertia of the drive, actuator, and control components in a hydraulic system, coupled with the fluidity and viscosity of the hydraulic oil itself, generalized inertial support energy is difficult to characterize. Therefore, this paper only considers energy changes at the output end of the hydraulic system, using the fluctuations in the hydraulic cylinder's output energy as the generalized inertial support energy. The specific expression of generalized inertial support energy in a hydraulic system is as follows:

[0164] E h =(P1-P2)·S·L

[0165] Where S is the effective area of ​​the hydraulic cylinder, P1 is the cylinder pressure before the system disturbance, P2 is the cylinder pressure after the system disturbance, and L is the effective working distance during the disturbance.

[0166] (3) Mechanical system

[0167] In mechanical systems, generalized inertia support energy primarily derives from the kinetic energy of moving parts. When the external excitation of a mechanical system changes, its generalized inertia support energy manifests as a change in kinetic energy. Because mechanical systems lack electrical current, hydraulic oil, or other fluid media, the quantification of generalized inertia can be directly applied to Newton's laws. The specific calculation formula is as follows:

[0168]

[0169] m is the mass of the moving part, I1 is the velocity of the moving part before the system disturbance, and I2 is the velocity of the moving part after the system disturbance; F is the thrust of the mechanical system, and x is the displacement of the mechanical system.

[0170] Through the above analysis, the present invention preliminarily determined the expression of generalized inertial support energy in electrical, hydraulic, and mechanical systems. However, due to the coupling effect of electrical, hydraulic, and mechanical systems, the inertia of construction equipment cannot be simply regarded as the sum of the three systems. In addition, in the process of calculating the generalized inertial support energy of electrical, hydraulic, and mechanical systems, many assumptions are introduced, resulting in a certain deviation between the calculation of generalized inertial support energy and actual conditions. To address these problems, generalized inertial relaxation factors λ1, λ2, and λ3 are introduced to accurately express the generalized inertial support energy of construction equipment, as shown below:

[0171]

[0172] Where, E e is the generalized inertia support energy of the electrical system, E h is the generalized inertia support energy of the hydraulic system, E m The generalized inertia support energy for the mechanical system. Because the energy transfer in electrical, hydraulic, and mechanical systems is linear, the inertia product phenomenon does not need to be considered. The coupling between the electrical, hydraulic, and mechanical systems is represented by relaxation factors, where the generalized inertia support energy of one system can enhance or weaken the inertia effect of another system.

[0173] Because a precise expression for inertia cannot be obtained through theoretical analysis, this paper introduces a data-driven approach. By establishing a proxy model for variables and the generalized inertia support energy, combined with an optimization algorithm, the decoupling of the relaxation factors λ1, λ2, and λ3 is achieved. Combining physical models with a data-driven approach allows for precise measurement of the generalized inertia support energy.

[0174] Step 3: Measurement of generalized inertness;

[0175] Because the contributions of electrical, hydraulic, and mechanical systems to generalized inertia cannot be precisely quantified in theoretical analysis, theoretical derivation alone cannot accurately derive the generalized inertia support energy of the entire device. Data-driven models, as black-box models, can construct response relationships between inputs and outputs, enabling parameter prediction and estimation, even when the operating mechanisms of these parameters are unknown. Therefore, this section introduces a deep learning model to construct proxy relationships between key electrical, hydraulic, and mechanical parameters and the generalized inertia support energy of the device.

[0176] In order to predict the generalized inertial support energy, an encoder-decoder neural network model integrating compressed excitation and shared weight strategy is designed. The model consists of an encoder and a decoder, such as Figure 3 As shown in Figure 2, the encoder uses convolutional layers, attention layers, a squeeze-excitation module, and dense layers to extract high-level features of the input parameters. Both the convolutional and attention layers use a standard network architecture. The squeeze-excitation module focuses on squeezing and excitation of the extracted features. Squeezing primarily involves global average pooling of the data, while excitation involves reconstructing features through full connectivity and activation functions. The squeeze-excitation calculation process is as follows:

[0177]

[0178] s=FC(z)

[0179] y=σ(s)⊙a

[0180] where a is the input feature map, z is the global feature vector obtained by applying global average pooling to x, FC represents a fully connected layer, s is the learning vector that models channel dependencies, σ() is the sigmoid activation function, and ⊙ represents the dot product.

[0181] The decoding path is essentially the same as the encoding path. The decoder network uses this jointly defined feature as input. The decoder outputs the predicted generalized lazy support energy. Sharing weights between the encoder and decoder ensures efficient and consistent training. The activation function for the shared weights is ReLU. The model is trained using supervised learning and converges by minimizing the loss function. The loss function is the mean squared error between the actual and predicted samples, as shown below:

[0182]

[0183] Where n is the number of training samples, L is the prediction error, θ is the network hyperparameter, f1(z) is the mapping function of the decoder, z is the feature vector, x i and x' ,i is the input feature.

[0184] To verify the model's accuracy, the output error of the prediction model is primarily measured using model evaluation metrics. Common evaluation metrics include the coefficient of determination (R²), root mean square error (RMSE), mean absolute error (MAE), and D² absolute error score (D²). For RMSE and MAE, smaller values ​​indicate higher model prediction accuracy. The closer R² and D² are to 1, the higher the model's prediction accuracy.

[0185] In order to further ensure the accuracy of the prediction model, the present invention has carried out a hyperparameter comparison of the model. The learning rate, number of iterations, optimizer and loss function are the core hyperparameters of the model. The comparison results are as follows: Figure 4 and Figure 5 As shown in the figure. By comparing different hyperparameters, the present invention ultimately selected a learning rate of 0.005 and 5000 iterations. To minimize model training error, the model ultimately selected the Rprop optimizer and the MSELoss loss function. All hyperparameter comparisons were averaged after 10 runs to effectively avoid parameter errors.

[0186] The results show that the proposed construction equipment inertia measurement model can accurately predict the generalized inertia support energy with an accuracy higher than 98%. Figure 6 As shown in Table 2, the R², RMSE, MAE, and D² for the training set were 0.994, 12.673, 6.384, and 0.976, respectively, while those for the test set were 0.987, 23.796, 14.29, and 0.957, respectively. The high R² value indicates that the model achieved good prediction results on both the training and test sets; the high D² value indicates that all model features were fully utilized. The R² and D² errors for the training and test sets were 0.7% and 1.95%, respectively, indicating that the model was not overfitted. Although the RMSE and MAE obtained by the prediction model were large, this was due to the magnitude of the generalized inert support energy data. The (RMSE / Mean)% for the training and test sets were 0.11% and 0.21%, respectively, indicating that the model's prediction bias was small compared to the initial data. The (MAE / Mean)% showed similar results, further demonstrating that the method was highly accurate for most samples. In summary, the encoder-decoder neural network model combined with compressed excitation and shared weight strategy can fully utilize the features in the training data and achieve high-precision prediction of generalized inert support energy.

[0187] For different construction scenarios, the encoder-decoder neural network model has high robustness and stability, and the model's prediction deviation is less than 2%. Figure 7The prediction results of the encoder-decoder neural network model in five scenarios are shown, sampled according to different construction intervals. The R2 of the five scenarios are 0.976, 0.982, 0.989, 0.988, and 0.991, respectively. Although the operating procedures, energy consumption, and operating environment of the five scenarios are different, the prediction deviations are all less than about 2% (0.976~0.991). As the operating power increases, the prediction error of the model does not increase or decrease significantly, indicating that the model is applicable to each operating range and has high robustness and stability. The results show that the variables selected by the present invention can fully characterize the inertia of the equipment and will not fail due to changes in the operating scenario. Although the data-driven model has good prediction effect and generalization ability for the generalized inertia support energy, it is unable to obtain the contribution of each subsystem to the inertia effect. Therefore, it is very necessary to combine the optimization model to perform decoupling analysis of the electrical, hydraulic, and mechanical systems.

[0188] Table 2 Model accuracy on the training and test sets.

[0189]

[0190]

[0191] Step 4: Decoupling analysis of generalized laziness

[0192] After completing proxy models for the generalized inertial support energy of electrical, hydraulic, and mechanical systems, the present invention combines data-driven models with physical models to perform decoupling analysis. Because the theoretical value of the generalized inertial support energy is generally consistent with the predictions of the proxy models, the decoupling problem of the generalized inertial relaxation factors λ1, λ2, and λ3 can be transformed into an optimization problem. By designing the corresponding optimization objectives and constraints, λ1, λ2, and λ3 are obtained. The optimization model is as follows:

[0193]

[0194] Variables:λ1,λ2,λ3

[0195] Constraint:

[0196] In order to ensure that the generalized inertia relaxation factor can meet the operating scenarios, the present invention designs the optimal target as the square difference between the theoretical value and the estimated value. The constraint equation is determined by the law of energy transfer, that is, hydraulic energy comes from electrical energy, and mechanical energy comes from hydraulic energy, and there is loss in the energy transfer process. It is worth mentioning that for different data sets (operating scenarios), the generalized inertia relaxation factor is often different. This is in line with the actual characteristics of construction equipment, because the inertia of electrical, hydraulic and mechanical systems in different operating scenarios is often different.

[0197] In order to obtain the optimal solution for the generalized inertia relaxation factor, the optimal foraging algorithm is introduced to solve the problem. The original optimal foraging theory was proposed by ecologists and zoologists to explain the resource utilization and eating patterns of animals. The optimal foraging theory states that animals tend to prey on prey that has the highest net energy intake per unit foraging time. Inspired by the optimal foraging theory, the optimal foraging algorithm was designed and applied to optimization problems. According to the optimal foraging theory, the optimization process can be viewed as an animal foraging in many plots until it finds the best plot that best maximizes its net energy intake rate. After finding the ideal plot, the animal uses the best prey model to select its position in the plot. Once the final ideal position is determined, the global optimization problem has its optimal solution. Each plot within the animal's foraging habitat can be conceptualized as a neighborhood of a local optimal solution to a function in the constraint space, or as a neighborhood of a global optimal solution. The pseudo code of the optimal foraging algorithm is:

[0198]

[0199] The proposed optimization algorithm achieves efficient decoupling of electrical, hydraulic and mechanical relaxation factors, with a decoupling accuracy greater than 99.8% and a decoupling efficiency of less than 60s. Figure 8 As shown in the figure, the final optimization target of the optimal foraging algorithm is 0.0019 (0.19%), and the algorithm converges in 38 cycles (58.4s), indicating that the physical equations and the agent model can accurately match. The optimal variables of the optimization model are [1.297, 0.970, 1.289], that is, the generalized inertia relaxation factors of the electrical, hydraulic and mechanical systems are 1.297, 0.970, and 1.289, respectively. The obtained generalized inertia support energy equation is Et = 1.297E. + 0.97Eh + 1.289Em, which is consistent with the actual situation. Due to E e Since the calculations for the electronic control components are ignored, the electrical system has the largest relaxation coefficient. In the hydraulic system, the energy storage properties of the hydraulic oil overcome the inertia phenomenon to a certain extent. Due to the lack of buffering, the inertia contribution of the mechanical system is also significant. The results show that the roadheader has a significant inertia effect within the mechanical system, which can be mitigated to some extent by the presence of the hydraulic system. By decoupling the system's generalized inertia support energy, the sensitivity of different systems to external responses can be clarified, facilitating the design and improvement of subsequent control strategies.

[0200] It should be noted that the embodiments of the present invention can be implemented by hardware, software, or a combination of software and hardware. The hardware portion can be implemented using dedicated logic; the software portion can be stored in a memory and executed by an appropriate instruction execution system, such as a microprocessor or dedicated design hardware. Those skilled in the art will appreciate that the above-mentioned devices and methods can be implemented using computer-executable instructions and / or contained in processor control code, for example, such as a carrier medium such as a disk, CD or DVDROM, a programmable memory such as a read-only memory (firmware), or a data carrier such as an optical or electronic signal carrier. Such code is provided on a carrier medium such as a disk, CD or DVDROM, a programmable memory such as a read-only memory (firmware), or a data carrier such as an optical or electronic signal carrier. The device and its modules of the present invention can be implemented by hardware circuits such as very large-scale integrated circuits or gate arrays, semiconductors such as logic chips, transistors, or programmable hardware devices such as field programmable gate arrays, programmable logic devices, etc., or can be implemented by software executed by various types of processors, or can be implemented by a combination of the above-mentioned hardware circuits and software, such as firmware.

[0201] This paper proposes an energy-based decoupling framework for inertia measurement. Based on the form of energy transfer in construction equipment, a physical expression method for the generalized inertia support energy of electrical, hydraulic, and mechanical systems is determined. A generalized inertia relaxation factor is used to decompose the inertia contribution of different systems. To accurately predict the generalized inertia support energy, an encoder-decoder neural network integrating compressed excitation and shared weight strategies is introduced. By combining the physics-driven model with the data-driven model, an inertia decoupling analysis model is established, and the generalized inertia relaxation factor is quantified. Experiments verify the reliability of this method.

[0202] In view of the case study, the main research results are summarized as follows. (1) The constructed data-driven measurement model of generalized inert support energy has high accuracy, and the R2, RMSE, MAE and D2 of the test set are 0.987, 23.796, 14.29 and 0.957 respectively. (2) The decoupling optimization model of generalized inert support energy has good efficiency and effect, and the optimal solution of the model is 0.0018, and the optimization time is 58.4s. (3) Through the decoupling optimization model, the theoretical expression of generalized inert support energy is, which reflects the degree of inertia contribution of electrical, hydraulic and mechanical systems. (4) The theoretical expression of generalized inert support energy is verified by actual engineering samples, and the error of 81.94% of the samples is less than 5%.

[0203] The present invention is verified by an on-site tunnel boring machine construction example. The error between the calculated value and the actual value on site is as follows: Figure 9 As shown. From the absolute error ( Figure 9-a, b), the calculation error of 92.18% of the samples in the inert energy calculation equation is less than 20kw, and the calculation error of 67.26% of the samples is less than 5kw. Since the roadheader is a device with high energy consumption, the present invention further provides the relative error ( Figure 9 -c, d) to further quantify whether the 20kW error is within an acceptable range. In the equation calculation results, 93.98% of the samples had a relative error of less than 20%, and 81.94% of the samples had a relative error of less than 5%, indicating that the generalized inertia measurement method fitted in this paper has high reliability, can truly reflect the inertia of the equipment subsystem, and has reliable application prospects.

[0204] The above description is only a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any modifications, equivalent substitutions and improvements made by any technician familiar with this technical field within the technical scope disclosed by the present invention and within the spirit and principles of the present invention should be covered by the scope of protection of the present invention.

Claims

1. A method for precise measurement and decoupling of generalized inertia of complex engineering equipment, characterized by: The following steps are involved: Step 1: Decomposition and summary of electrical, hydraulic and mechanical systems in building equipment; Analyze the basic structure of different construction equipment, determine the energy transfer paths between different systems, identify the factors affecting the inertia of the equipment and the main sources, and identify the common inertia forms in complex engineering equipment; Step 2: Theoretical analysis and derivation of generalized inertia; Determine the generalized inertia expressions for the electrical, hydraulic, and mechanical parts, derive formulas to construct a calculation method for the generalized inertia support energy of the entire machine, introduce a generalized inertia relaxation factor, and integrate the inertia of each subsystem to form a theoretical expression of the inertia of the entire machine; Step 3: Measurement of generalized inertia; A data-driven proxy model is constructed for the parameter of the generalized inertia support energy of the entire device. A deep neural network framework for encoding and decoding is designed to accurately measure the device inertia and achieve data-based generalized inertia characterization. Step 4: Decoupling analysis of generalized inertia; A parameter decoupling model is constructed by combining the theoretical analysis model of inertia with the data-driven model. The optimization goal is to make the measurement results of the two consistent, thus achieving the decoupling of the inertia relaxation factors of the electrical, hydraulic, and mechanical parts, and ensuring the accurate theoretical expression of the inertia of the entire machine. The generalized inertia support energy of the electric drive system is regarded as the main inertia source of the electric drive system and is expressed as: Where J is the moment of inertia, ω1 is the mechanical angular velocity before the system disturbance, ω2 is the mechanical angular velocity after the system disturbance, T is the output torque, and t is the time. The fluctuation of the hydraulic cylinder output energy in the hydraulic system is regarded as the generalized inertial support energy. The specific expression is as follows: E h =(P1-P2)·S·L Where S is the effective area of ​​the hydraulic cylinder, P1 is the cylinder pressure before the system disturbance, P2 is the cylinder pressure after the system disturbance, and L is the effective working distance during the disturbance. The generalized inertial support energy of a mechanical system is expressed as a change in kinetic energy. The specific calculation formula is as follows: Where m is the mass of the moving part, v1 is the velocity of the moving part before the system disturbance, v2 is the velocity of the moving part after the system disturbance, F is the thrust of the mechanical system, and x is the displacement of the mechanical system. The generalized inertial relaxation factors λ1, λ2, and λ3 are introduced to accurately express the generalized inertial support energy of construction equipment, as shown below: Where, E e is the generalized inertia support energy of the electrical system, E h is the generalized inertia support energy of the hydraulic system, E m It is the generalized inertia support energy of the mechanical system.

2. The method for precise measurement and decoupling of generalized inertia of complex engineering equipment according to claim 1 is characterized in that: The electrical system: The generalized inertia support energy of the electrical system consists of two parts, namely the electric drive part and the electronic control part. The electric drive part is mainly composed of various motors. For a running motor, the inertia is expressed by the degree of resistance to speed changes. The rotational inertia of the motor is expressed as: J = ∫r 2 dm where r is the radius of rotation; m is the mass of the rigid body; the unit of moment of inertia J is kgm 2 , for the motor, the moment of inertia is constant; When an unbalanced disturbance occurs in the motor, the inertia is manifested as fluctuations in the motor output energy. Therefore, the generalized inertia support energy of the electric drive part is expressed as: Where J is the moment of inertia, ω1 is the mechanical angular velocity before the system disturbance, ω2 is the mechanical angular velocity after the system disturbance, T is the output torque, and t is time. The generalized inertia support energy of the motor is expressed as the energy change before and after the system disturbance, that is, the energy of the system to overcome the inertia and do work. In the electronic control system, the generalized inertial support energy is expressed by the response changes to different electrical signals, and its expression is: AND e2 =V∫I1dt-∫I2dt Where V is the voltage of the control circuit, I1 is the current of the control signal before the system disturbance, and I2 is the current of the control signal after the system disturbance. Compared with the electric drive system, the generalized inertia support energy of the electric control system is much smaller. The generalized inertia support energy of the electric drive system is regarded as the main inertia source of the electric transmission system.

3. The method for precise measurement and decoupling of generalized inertia of complex engineering equipment according to claim 2 is characterized in that: The hydraulic system: Due to the inertia of the driving, executing, and controlling components in the hydraulic system, as well as the fluidity and viscosity of the hydraulic oil itself, the inertia within the hydraulic system can be reflected at the output end. Therefore, the fluctuation of the hydraulic cylinder output energy is regarded as the generalized inertia support energy. The specific expression of the generalized inertia support energy in the hydraulic system is as follows: E h =(P1-P2)·S·L Where S is the effective area of ​​the hydraulic cylinder, P1 is the cylinder pressure before the system disturbance, P2 is the cylinder pressure after the system disturbance, and L is the effective working distance during the disturbance.

4. The method for precise measurement and decoupling of generalized inertia of complex engineering equipment according to claim 3 is characterized in that: The mechanical system: In mechanical systems, generalized inertia support energy primarily comes from the kinetic energy of moving parts. When the external excitation of a mechanical system changes, its generalized inertia support energy manifests as a change in kinetic energy. Since there is no electric current, hydraulic oil, or other fluid medium in a mechanical system, the quantification of its generalized inertia can be directly referenced by Newton's laws. The specific calculation formula is as follows: m is the mass of the moving part, v1 is the velocity of the moving part before the system disturbance, v2 is the velocity of the moving part after the system disturbance, F is the thrust of the mechanical system, and x is the displacement of the mechanical system; The generalized inertial relaxation factors λ1, λ2, and λ3 are introduced to accurately express the generalized inertial support energy of construction equipment, as shown below: Where, E e is the generalized inertia support energy of the electrical system, E h is the generalized inertia support energy of the hydraulic system, E m It is the generalized inertia support energy of the mechanical system.

5. The method for precise measurement and decoupling of generalized inertia of complex engineering equipment according to claim 1 is characterized in that: An encoder-decoder neural network model with integrated compression excitation and shared weight strategies is used. The model consists of an encoder and a decoder. The encoder uses convolutional layers, attention layers, compression excitation modules, and dense layers to extract high-level features of the input parameters. Both the convolutional and attention layers use standard network structures. The compression excitation module focuses on squeezing and excitation of the extracted features. Squeezing mainly performs global average pooling on the data, while excitation reconstructs the features through full connectivity and activation functions. The calculation process of compression excitation is as follows: s=FC(z) y=σ(s)⊙a Where a is the input feature map, z is the global feature vector obtained by applying global average pooling to x, FC represents a fully connected layer, s is the learning vector that models channel dependencies, σ() is the sigmoid activation function, and ⊙ represents the dot product. The decoding path is basically the same as the encoding path; the decoder network uses the features of the decoding path and the encoding path as input; The decoder outputs the predicted generalized inert support energy. By sharing weights between the encoder and decoder, the training process is efficient and consistent. The activation function for the shared weights is ReLU. The model is trained in a supervised learning manner and converges by minimizing the loss function. The loss function is the mean squared error between the actual sample and the predicted sample, as shown below: Where n is the number of training samples, L is the prediction error; θ is the network hyperparameter, f1(z) is the mapping function of the decoder, z is the feature vector, x i and x ',i is the input feature.

6. The method for precise measurement and decoupling of generalized inertia of complex engineering equipment according to claim 1, characterized in that: Decoupling analysis of the generalized laziness: By designing the corresponding optimization objectives and constraints, we can obtain λ1, λ2, and λ3. The optimization model is as follows: Variables: λ1, λ2, λ3 Constraints:

7. A generalized inertia precision measurement and decoupling system for complex engineering equipment that implements the method according to any one of claims 1 to 6, characterized in that: include: Decomposition and Induction Module: This module is used to analyze the basic structure of construction equipment, determine the energy transfer paths between different systems, and identify the inertia-influencing factors and main sources of the electrical, hydraulic, and mechanical systems in the equipment, as well as the common inertia forms in complex engineering equipment. Theoretical analysis and derivation module: used to determine the generalized inertia expressions of the electrical, hydraulic, and mechanical parts, construct a generalized inertia support energy calculation method for the entire machine, and introduce a generalized inertia relaxation factor to integrate the inertia of each subsystem to form a theoretical expression of the inertia of the entire machine; Measurement module: Builds a data-driven proxy model for the generalized inertia support energy of the entire device and uses an encoding and decoding deep neural network framework to accurately measure device inertia, thereby achieving data-based characterization of generalized inertia. Decoupling Analysis Module: This module combines the theoretical analysis model of inertia with the data-driven model to construct a parameter decoupling model. The module designs optimization targets to ensure consistency between the two measurement results, decoupling the inertia relaxation factors of the electrical, hydraulic, and mechanical components, and ensuring accurate theoretical expression of the inertia of the entire machine.

8. The generalized inertia precision measurement and decoupling system for complex engineering equipment according to claim 7, characterized in that: include: Electric drive analysis module: used to analyze the motor's moment of inertia and the fluctuation of the motor's output energy when an unbalanced disturbance occurs, thereby determining the generalized inertial support energy of the electric drive part; Electronic control part analysis module: used to analyze the response changes of the control circuit to different electrical signals to determine the generalized inertial support energy of the electronic control system; Integration module: used to integrate the generalized inertial support energy of the electric drive part and the electronic control part, serving as the main inertial source of the electrical system.

9. The generalized inertia precision measurement and decoupling system for complex engineering equipment according to claim 7, characterized in that: include: Hydraulic cylinder output analysis module: used to analyze the output energy fluctuations of the hydraulic cylinder before and after system disturbances to determine the generalized inertial support energy of the hydraulic system; Generalized inertia quantification module: used to quantify the generalized inertia support energy of the hydraulic system based on the fluctuation of the hydraulic cylinder output energy.

10. The generalized inertia precision measurement and decoupling system for complex engineering equipment according to claim 7, characterized in that: include: Kinetic Energy Analysis Module: used to analyze the changes in the kinetic energy of a mechanical system when the external excitation changes, so as to determine the generalized inertial support energy of the mechanical system; Generalized Inertia Quantification Module: used to quantify the generalized inertia support energy of a mechanical system based on the change in kinetic energy.

Images