A large-diameter box culvert template self-adaptive installation and dismounting adjusting method and system

Abstract

This invention discloses an adaptive installation and dismantling adjustment method and system for large-diameter box culvert formwork, belonging to the field of intelligent control technology in civil engineering construction. The method includes: acquiring pressure data through hydraulic preloading to solve for the initial demolding load; when the load exceeds a threshold, activating an asymmetric vibrator and configuring a phase angle deviation to excite traveling wave modes on the formwork panel to break the adsorption stress; rapidly pressurizing and depressurizing the retraction cylinder through a pressure regulating actuator to cause the formwork to reciprocate; after a gap is formed, dynamically adjusting the excitation force amplitude and pulsation frequency based on the instantaneous demolding speed fed back by a displacement sensor using a closed-loop adjustment algorithm until the formwork is completely removed. The system includes a sensing layer, a decision layer, and an execution layer. This invention can proactively release the strong adsorption stress between the formwork and concrete, avoid demolding damage, and improve the intelligence level and safety of large-diameter box culvert construction.

Description

Technical Field

[0001] This application belongs to the field of intelligent control technology for civil engineering construction, specifically relating to an adaptive installation and dismantling adjustment method and system for large-diameter box culvert formwork. Background Technology

[0002] With the deepening development of underground space in modern infrastructure construction, large-diameter box culverts, as a core component of integrated utility tunnels and large-scale water diversion projects, have a decisive impact on the durability of the overall structure due to their construction quality. In the standardized operations of box culvert construction, the formwork system, as a key process equipment for concrete forming, directly affects construction costs and on-site safety through its installation accuracy and dismantling efficiency. As construction technology transitions from mechanization to intelligentization, more stringent requirements are placed on the precise adaptation and condition adjustment of large formwork under complex working conditions.

[0003] Among these, the adaptive dismantling and posture adjustment of large-diameter box culvert formwork is a core element in ensuring concrete molding quality and equipment recyclability. This technology aims to achieve smooth formwork removal after molding by precisely controlling the interface force between the formwork and concrete, thereby ensuring the appearance quality of the box culvert's inner wall while avoiding damage to the supporting structure due to uneven stress. Given the engineering characteristics of large diameter, high self-weight, and large contact area, real-time monitoring and dynamic adjustment of the dismantling process has become an important direction for optimizing construction techniques.

[0004] However, traditional techniques for dismantling large-diameter box culvert formwork rely primarily on the rigid pulling action of heavy-duty hydraulic cylinders, which struggles to overcome the intense adsorption stress generated by the large contact area between the formwork and concrete. This forced demolding method easily generates significant negative pressure at the moment of demolding, leading to severe defects such as pitting, cracking, and localized spalling on the concrete surface. Furthermore, because the pulling force cannot be dynamically distributed based on interfacial resistance, the persistent asymmetric stress causes cumulative fatigue deformation of the formwork frame, significantly reducing its lifespan. In addition, existing dismantling schemes lack effective feedback and adjustment mechanisms, failing to proactively release stress upon detecting abnormal resistance, resulting in insufficient intelligence and safety hazards in the construction process. Therefore, an adaptive installation and dismantling adjustment scheme for large-diameter box culvert formwork is desired. Summary of the Invention

[0005] The purpose of this invention is to provide an adaptive installation and dismantling adjustment method and system for large-diameter box culvert templates, which can effectively solve the problems mentioned in the background art.

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

[0007] A method for adaptive installation and dismantling adjustment of formwork for large-diameter box culverts includes the following specific steps:

[0008] Step S1: The hydraulic system performs a small preload on the retracting cylinder, and the pressure data is collected by the pressure transmitter; the central processing unit filters the collected pressure data and, in combination with the pre-stored mechanical efficiency correction coefficient and the total contact area between the template panel and the concrete, calculates the initial demolding load.

[0009] Step S2: Compare the initial demolding load with the set pressure threshold. When it is greater than or equal to the threshold, start the asymmetric vibrator arranged on the reinforcing rib node on the back of the template. By configuring the vibration phase angle deviation between adjacent vibrators, the traveling wave mode is excited on the template panel.

[0010] Step S3: While maintaining the operation of the vibrator, the hydraulic oil in the piston chamber of the retracting cylinder is rapidly pressurized and depressurized by the pressure regulating actuator, so that the pressure amplitude alternates between the peak value and the valley value, driving the template panel to reciprocate relative to the concrete interface.

[0011] Step S4: After a gap is formed between the edge of the template and the concrete, the current position value of the template detected by the displacement sensor is obtained and the instantaneous demolding speed is calculated. The central processing unit runs a closed-loop adjustment algorithm to maintain the instantaneous demolding speed near the preset standard demolding speed, and dynamically adjusts the excitation force amplitude and pulsation frequency according to the speed deviation until the template is completely removed from the working surface.

[0012] Furthermore, the specific configuration of the asymmetric exciter in step S2 is as follows: each asymmetric exciter integrates a double eccentric block structure for adjusting the excitation force amplitude and a micro servo motor for independently programmed control of the drive phase angle.

[0013] When the central processing unit generates multi-channel exciter drive pulses, it applies a phase angle deviation of 15 to 45 degrees between two or more spatially adjacent groups of exciters to break the overall resonance trend of the template panel and form a traveling wave mode with non-uniform spatial distribution and fluctuating time series characteristics.

[0014] At the same time, the activation command issued by the central processing unit activates the exciters in a fan-shaped sequence that spreads from the geometric center of the template to the surrounding edges.

[0015] Furthermore, in step S3, the pressure regulating actuator is specifically a high-frequency proportional relief valve or a high-speed switching solenoid valve with a step response time of less than 10 milliseconds.

[0016] The central processing unit generates a pulse width modulation signal with a variable duty cycle or an analog signal with continuously varying amplitude according to a preset pulse control strategy to drive the pressure regulating actuator.

[0017] The frequency of the high-speed alternating pressure amplitude is determined by the central processing unit through frequency sweeping to identify the low-order natural resonant frequency of the template reinforcing rib frame, and the working range is locked in the 15Hz to 35Hz frequency band that avoids the low-order natural resonant frequency. The waveform of the pressure amplitude change is a sine wave.

[0018] Furthermore, the closed-loop control algorithm in step S4 is a fuzzy PID control architecture;

[0019] The central processing unit first calculates the deviation e between the instantaneous demolding speed and the standard demolding speed, as well as the rate of change ec of the deviation. Then, it multiplies the calculated e and ec by a preset quantization factor and maps them to a fuzzy subset on the fuzzy domain.

[0020] The central processing unit performs rule matching and inference synthesis based on the pre-stored fuzzy control rule table, and uses the centroid method to perform defuzzification operation on the fuzzy set obtained by inference to obtain the real-time incremental correction values ​​of PID controller parameters Kp, Ki, and Kd. Then, it calculates the control quantities used to adjust the amplitude of the excitation force of the asymmetric exciter, the hydraulic pulsation frequency, and the target pressure of the proportional relief valve.

[0021] Furthermore, the closed-loop adjustment algorithm in step S4 is a model predictive control architecture;

[0022] The central processing unit internally maintains a linear time-invariant prediction model with template displacement and instantaneous velocity as state vectors and excitation force amplitude instructions and hydraulic pulsation frequency instructions as input vectors;

[0023] Within each control cycle, the central processing unit (CPU) starts from the current measured state and performs rolling predictions of the output trajectory within a finite time step in the future. It then solves for the optimal control sequence at the current moment by minimizing a quadratic cost function that minimizes the tracking error between the predicted speed and the set speed.

[0024] Furthermore, during the execution of step S4, a real-time monitoring mechanism for the uniformity of force on the retracting cylinder is also included: the central processing unit polls and scans the load pressure values ​​obtained by the pressure sensor embedded in the piston rod head of each retracting cylinder at a fixed period, and calculates the arithmetic mean of the load pressure of all cylinders.

[0025] When the load value of a certain cylinder is detected to deviate from the arithmetic mean by more than 15% of the average load, the off-center load compensation subroutine is activated. By adjusting the opening of the electromagnetic proportional flow valve independently associated with each retraction cylinder, the flow input is increased on the side with less force, and the flow is restricted on the side with more force.

[0026] Furthermore, the method also includes a demolding strategy self-optimization step: organizing all-dimensional parameters, including pressure signal, displacement signal, vibration frequency, pulsation amplitude, ambient temperature, and concrete age, into a structured demolding condition package and uploading it to a cloud database.

[0027] The regression analysis algorithm built into the central processing unit takes the peak value of the initial demolding load for each demolding operation, the time taken to reach the stable shrinkage stage, and the surface quality score of the inner wall of the box culvert as optimization target variables, and takes the ambient temperature, concrete age, and water-cement ratio as input feature variables. It fits the functional relationship between the input features and the optimal process parameters through multiple linear regression or support vector regression to automatically optimize the subsequent demolding strategy.

[0028] Furthermore, the method also includes multi-level safety protection logic: the central processing unit is set with multi-level pressure protection thresholds. When the hydraulic system pressure abnormally surges to 110% of the system's rated pressure or the displacement sensor detects no displacement change within 5 consecutive seconds, the central processing unit immediately cuts off the main oil circuit solenoid valve to execute an emergency stop and sends an audible and visual alarm signal to the operating terminal. At the same time, it automatically locks the hydraulic cylinder position and switches to manual intervention mode.

[0029] After the template completely exits the working surface and touches the rear limit switch, the system automatically performs a pressure relief operation to reduce the hydraulic circuit pressure to a safe value.

[0030] Furthermore, the method is also applicable to the template installation and adjustment stage: during the process of advancing the template to the pouring position, the central processing unit drives the retraction cylinder to execute fine-tuning instructions, and the displacement sensor provides real-time feedback on the relative distance between the template and the reference surface;

[0031] When the relative distance enters the fine-tuning range of 50 mm, the system automatically switches to slow approach mode, reducing the advancing speed to 0.5 mm / s;

[0032] Contact-type limit switches placed at the four corners of the template monitor the tightness of the template's fit. After installation, the system automatically collects the current displacement coordinates as the zero point and stores them in memory as the reference for the next demolding calculation.

[0033] A large-diameter box culvert formwork adaptive installation and dismantling adjustment system, including

[0034] The sensing layer includes a diffused silicon pressure transmitter installed at the high-pressure oil inlet port of the retraction cylinder, displacement sensors arranged between the four corners and the midpoint of the long side of the template frame and the fixed support base, and a pressure sensor embedded in the piston rod head of each retraction cylinder, which are used to collect pressure data, displacement data and load distribution data respectively.

[0035] The decision-making layer includes a central processing unit (CPU) and a programmable logic controller (PLC). The CPU performs filtering on the acquired data, calculates the initial demolding load, and compares the initial demolding load with a set pressure threshold to generate stress release, pulsating retraction, or closed-loop adjustment commands. The PLC establishes a data link with the variable frequency drive of the hydraulic pump station via an industrial real-time Ethernet protocol and receives and executes the commands from the CPU.

[0036] The execution layer includes an asymmetric vibrator located at the cross intersection of the reinforcing ribs on the back of the template structure, a high-frequency proportional relief valve or a high-speed switching solenoid valve installed in the hydraulic circuit, and an electromagnetic proportional flow valve independently associated with each retraction cylinder. The asymmetric vibrator is used to excite traveling wave modes according to the instructions of the decision layer. The high-frequency proportional relief valve or the high-speed switching solenoid valve is used to quickly pressurize and depressurize the hydraulic oil in the piston chamber of the retraction cylinder. The electromagnetic proportional flow valve is used to adjust the oil inlet of the corresponding cylinder during off-center load compensation.

[0037] In summary, this application includes at least one of the following beneficial technical effects:

[0038] 1. This invention utilizes the synergistic effect of a load prediction mechanism and a stress release triggering mechanism to effectively disrupt the vacuum adsorption stress between the formwork and concrete during the initial demolding stage by employing traveling waves generated by an asymmetric vibrator. This method significantly reduces instantaneous demolding resistance. Due to the active release of the adsorption force, the pull-out damage to the concrete surface caused by the negative pressure generated at the moment of demolding is completely eliminated, effectively preventing defects such as pitting, cracking, and localized spalling, thus ensuring that the appearance quality of the inner wall of large-diameter box culverts reaches excellent standards.

[0039] 2. This invention employs a pulsating retraction control and a multi-cylinder force balancing mechanism, overcoming the drawbacks of uneven force distribution and constant tension in traditional retraction cylinders. Through periodic pressure pulsation and real-time off-center load compensation, the rigid impact stress on the formwork frame is reduced. By avoiding cumulative fatigue deformation caused by asymmetric stress, the structural stability of the formwork frame is maintained over a long period, significantly increasing the turnover rate and recycling rate of large formwork, and reducing construction costs.

[0040] 3. This invention constructs a closed-loop control system based on dual feedback of pressure and displacement, achieving fully unmanned control of the entire process from load prediction and stress release to smooth retraction. The system can dynamically adjust the vibration frequency and pulsation amplitude according to the real-time feedback of the demolding speed, ensuring that the demolding process is always under optimal stress. Compared with the traditional mode that relies on manual judgment and experience, this invention greatly improves the intelligence and safety of construction, not only shortening the operation time of a single demolding operation, but also enhancing the level of refined management in large-scale engineering construction. Attached Figure Description

[0041] Figure 1 A schematic diagram of the overall technical solution for the adaptive installation and dismantling adjustment method of large-diameter box culvert templates;

[0042] Figure 2 A schematic diagram illustrating the core principle of interface stress release based on asymmetric excitation and pressure pulsation;

[0043] Figure 3 The flowchart shows the adaptive adjustment logic of the demolding state based on multi-parameter closed-loop feedback.

[0044] Figure 4 This diagram illustrates the multi-level interaction relationships and data flow between the perception layer, decision-making layer, and execution layer. Detailed Implementation

[0045] To make the objectives, technical solutions, and advantages of this invention clearer, the following description is provided in conjunction with the appendix. Figures 1 to 4 The present invention will be further described in detail below with reference to specific embodiments.

[0046] Firstly, in the adaptive installation and removal adjustment method for large-diameter box culvert templates, the core execution hardware architecture is built upon a deep coupling of an integrated hydraulic drive system, a multi-dimensional sensing sensor network, and a high-performance industrial programmable logic controller. The implementation process strictly follows the technical path of data perception, logical decision-making, precise execution, and closed-loop optimization, and is carried out according to the following steps:

[0047] Firstly, in step S1, the actual bond strength between the template panel and the hardened concrete interface is accurately calculated through micro-preloading of the hydraulic system and data fusion with a high-precision sensor network. This result will serve as the quantitative basis for determining whether to initiate subsequent strong demolding assistance measures, and its accuracy directly affects the timing of subsequent execution and parameter strength settings.

[0048] In step S101, the system first confirms the curing status of the culvert's concrete. When feedback from the external concrete maturity monitoring system or the prediction model based on the age-strength curve indicates that the concrete compressive strength has reached the preset demolding threshold of 75% to 85% of the design strength grade, the central processing unit immediately generates a start command and sends it to the programmable logic controller. The issuance of this start command signifies that the system has switched from a long-term standby monitoring mode to an active demolding control mode.

[0049] In step S102, upon receiving the start command, the programmable logic controller (PLC) immediately establishes a real-time data link with the variable frequency drive (VFD) of the hydraulic pump station via its onboard high-speed communication interface. The high-speed communication interface employs an industrial real-time Ethernet protocol, such as Profinet bus or EtherCAT bus. This communication mechanism ensures that all subsequent control commands and sensor feedback signals can be exchanged bidirectionally within milliseconds.

[0050] In step S103, after receiving the operation enable signal via the data link, the power unit inside the hydraulic pump station drives the main oil pump to start outputting high-pressure hydraulic oil at a pre-calibrated displacement. The hydraulic oil first flows through a multi-stage filtration device to filter out any possible solid particulate contaminants, and then enters the main oil supply circuit.

[0051] The electromagnetic proportional directional valve, located in the main oil supply circuit, receives analog current commands from the programmable logic controller (PLC). It precisely adjusts the valve spool opening to throttle the flow rate, allowing hydraulic oil to be smoothly injected into the piston chamber of the retracting cylinder at a predetermined flow rate. This predetermined flow rate typically ranges from 10 liters per minute to 50 liters per minute, with the specific value selected by the central processing unit based on the current culvert diameter and a preset initial loading rate curve.

[0052] This effect creates stable physical conditions for accurate readings of the pressure transmitter in the subsequent step S104.

[0053] In step S104, as hydraulic oil is continuously injected into the piston chamber of the retracting cylinder, the diffused silicon pressure transmitter installed at the high-pressure oil inlet port of each retracting cylinder begins to perform high-frequency dynamic monitoring of the instantaneous pressure fluctuations inside the oil circuit. The selected pressure transmitter has a signal response bandwidth of not less than 1 kHz, and its upper limit of measurement range is adapted to the system's rated working pressure of 31.5 MPa.

[0054] Throughout its full range, the transmitter's linearity error is strictly controlled to within 0.1%. The pressure transmitter converts the sensed mechanical pressure value into a standard industrial analog electrical signal in real time, such as a current signal of 4 mA to 20 mA, or a voltage signal of 0 V to 10 V. This analog electrical signal is transmitted via shielded twisted-pair cable to a high-speed analog input module on the programmable logic controller rack.

[0055] In step S105, the analog-to-digital converter built into the high-speed analog input module discretizes and quantizes the received continuous analog signal at a sampling frequency of not less than 1000 Hz, generating a time-varying digital pressure sequence. This digital sequence is periodically and non-blockingly written to the internal random access memory register mapping area of ​​the central processing unit via the controller backplane bus.

[0056] In step S106, after acquiring the pressure sequence of multiple consecutive sampling periods, the central processing unit starts the signal preprocessing logic. First, the median filtering algorithm is executed. The core of the algorithm is to sort the odd number of sampling points in the sliding window and take the median value to remove occasional spike pulses caused by electromagnetic interference or instantaneous valve core movement.

[0057] Subsequently, a moving average filtering algorithm is used to perform secondary smoothing on the median-filtered sequence to further suppress the inherent pulsation noise of long-distance hydraulic pipeline transmission and the residual transient impact during pump source variable mechanism reversal. The pressure reading array processed by this composite filtering mechanism is recognized by the system as a steady-state effective value that can represent the current actual thrust level of the hydraulic cylinder.

[0058] Step S107: The central processing unit retrieves the geometric parameter configuration table of the retraction cylinders, which is pre-stored in the non-volatile memory. This configuration table contains at least the following two pieces of data: the effective force-bearing area of ​​the piston chamber of each retraction cylinder, in square meters; and the mechanical efficiency correction factor for each retraction cylinder, determined by factory bench testing or periodic in-situ no-load calibration.

[0059] Subsequently, the central processing unit calculates the initial demolding load at the current moment based on its built-in physical model. The formula used for the calculation is as follows:

[0060]

[0061] in, This represents the initial demolding load calculated from the output. Physically, it represents the sum of the molecular bonding force and vacuum adsorption force per unit area between the current formwork panel and the concrete interface, measured in Pascals. Representing the The measured steady-state pressure value at the oil inlet of the retraction cylinder after filtering in step S106 is expressed in Pascals. Representing the The mechanical efficiency correction coefficient for each retraction cylinder. This coefficient is dimensionless and its value is typically between 0.85 and 0.98. The purpose of introducing this coefficient is to deduct mechanical losses such as piston seal friction resistance and piston rod guide sleeve damping from the total thrust of the cylinder, so as to restore the actual net tensile force acting on the template panel.

[0062] This represents the total contact area between the formwork panel and the hardened concrete. This value is pre-calculated based on the box culvert's designed inner diameter, the length of a single section, and the sealing boundary conditions of the end grout seal, and is entered into the system parameter configuration table. The unit is square meters.

[0063] In step S108, the central processing unit processes the measured pressure values ​​of each hydraulic cylinder obtained in step S106. The corresponding mechanical efficiency correction coefficient retrieved in step S107 Multiply and sum the results, then divide by the total contact area. This allows us to calculate the initial demolding load at the current moment. .

[0064] This calculation process effectively removes the mechanical friction loss component within the hydraulic actuator through a weighted correction method, resulting in a higher final output load index. It can accurately and realistically characterize the adsorption strength resultant force between the template surface and the concrete interface.

[0065] Step S109: After completing the calculation of the initial demolding load, the central processing unit will process the calculated load... The value, along with the current timestamp, ambient temperature sensor readings, and independent pressure distribution data for each cylinder, is encapsulated into a structured data record.

[0066] This record is temporarily stored in the controller's local circular buffer. It serves two purposes: firstly, it allows for immediate retrieval by the threshold comparison logic in step S2; secondly, it is uploaded to the cloud-based process database via the industrial IoT gateway for subsequent self-optimization regression analysis of the demolding strategy. At this point, all sub-steps of the initial load monitoring and data fusion phase of step S1 have been completed.

[0067] In summary, step S1 utilizes mixed-signal processing technology and a physics-based efficiency correction algorithm to achieve a non-destructive quantitative assessment of the interfacial resistance of large formwork panels. This stage begins by confirming the physical conditions under which the concrete can be demolded. Through precise hydraulic flow control and a high-frequency pressure sensing network, the initial demolding load value, reflecting the true viscous state, is successfully obtained without inducing macroscopic displacement of the interface. This value completely eliminates the ambiguity of human experience-based judgment, providing the system with a rigid criterion for whether to enter the high-intensity demolding intervention process.

[0068] The next step is S2, after the initial demolding load is completed in step S1. After quantitative evaluation, the system's control logic transitions to an event-driven stress release phase. The core of this phase is that when the interfacial adhesion force exceeds the safety threshold for smooth pulling of the hydraulic system, the rigid demolding method of continuous pressurization is no longer relied upon. Instead, a mechanical vibration energy field is introduced to perform non-contact physical intervention on the bonding interface, aiming to actively dissipate harmful vacuum adhesion stress before the template undergoes macroscopic displacement.

[0069] In step S201, the central processing unit first reads the initial demolding load calculated in step S1 from the specified address of the internal register. Simultaneously, the central processing unit accesses the process parameter table stored in non-volatile memory and retrieves the set pressure threshold corresponding to the currently constructed box culvert. This sets the pressure threshold. It is not a fixed constant, but a reference benchmark dynamically generated by the system based on multiple input variables.

[0070] Step S202, to ensure that those skilled in the art can reproduce the set pressure threshold The generation process of is now described in detail, including its calculation basis and underlying logic. The dynamic generation depends on the function mapping of the following four types of parameters: First, the cross-sectional geometry of the box culvert, including the inner diameter span and the longitudinal length of a single pour, which is used to determine the basic magnitude of the contact area; Second, the water-cement ratio parameter in the concrete mix proportion, which directly affects the chemical bonding force and micropore structure of the cementitious material at the hardened interface; Third, the coverage model of the interface release agent, which is determined by the process parameters of the spraying operation and quantifies the actual film integrity of the release agent on the template panel in the form of area percentage; Fourth, the current real-time ambient temperature and humidity data, which is used to correct the influence of concrete surface moisture content and temperature difference deformation on the adsorption force.

[0071] For large-diameter box culvert structures with an inner span exceeding 8 meters, a pressure threshold is set. The specific value is determined as follows: the system uses the unit area adsorption force of concrete with the same mix proportion measured under laboratory conditions as a baseline, multiplied by the total contact area between the current formwork and the concrete. Furthermore, an additional pre-tightening safety factor with a value ranging from 1.2 to 1.5 is introduced on the basis of the resulting product.

[0072] The purpose of introducing this safety factor is to compensate for the deviation in the prediction of adsorption force caused by the fluctuation of the construction site environment and the dispersion of materials, and to ensure that the stress release mechanism can intervene in advance before interface damage occurs.

[0073] Step S203, the numerical comparator inside the central processing unit compares the initial demolding load. With set pressure threshold Execute real-time comparison logic. When When the system determines that the current interface viscosity is within the normal range, the template can be directly and smoothly removed under the quasi-static tension of the retraction cylinder. At this time, the process will skip steps S2 and S3 and directly enter the conventional smooth retraction control mode described in step S4.

[0074] when At that moment, the central processing unit (CPU) determines that the template panel is in a high-intensity viscous state, meaning that there is a strong adsorption field at the interface formed by vacuum negative pressure and molecular adhesion. Simply relying on the rigid pulling of the retracting hydraulic cylinder would risk damaging the surface quality of the concrete and causing fatigue deformation of the template frame. The moment this determination is confirmed, the CPU immediately activates the stress release protection logic.

[0075] In step S204, after the stress relief protection logic is triggered, the central processing unit sends a start command to multiple asymmetric vibrators located at key nodes of the reinforcing ribs on the back of the template via the synchronous pulse output module. These asymmetric vibrators are spatially arranged in an asymmetric array, with specific installation points selected at the cross intersection nodes of the longitudinal and transverse reinforcing ribs of the template.

[0076] The purpose of choosing this type of node as the excitation source is that the cross nodes of the reinforcing ribs have high structural stiffness and admittance characteristics, which can efficiently transmit the mechanical energy generated by the exciter to various areas of the metal panel along the reinforcing rib frame with a low attenuation rate, thereby achieving full coverage of vibration energy.

[0077] In step S205, each asymmetric exciter integrates a double eccentric block structure as the vibration source. By adjusting the rotation angle of the built-in micro servo motor, the relative angle between the two sets of eccentric blocks can be changed, thereby achieving stepless adjustment of the excitation force amplitude output by the exciter. Simultaneously, the drive phase angle of the micro servo motor can be independently programmed and controlled by a pulse sequence issued by the central processing unit, enabling each exciter to have independent amplitude and phase adjustment capabilities.

[0078] Step S206: The specific engineering implementation of the "asymmetric" characteristic lies in the differentiated configuration of the vibration start-up phase angle. When generating multi-channel exciter drive pulses, the central processing unit intentionally applies a phase angle deviation between two or more spatially adjacent exciters. This deviation value is set within a predetermined angle range of 15 to 45 degrees. This artificially introduced phase interference effect breaks the tendency of the template panel, as a continuous elastic body, to easily generate overall resonance at the same frequency under conventional excitation.

[0079] Instead, the template panel exhibits a traveling wave mode under forced vibration, which is spatially non-uniform and temporally fluctuating. From a physical perspective, the crests and troughs of the traveling wave continuously move across the panel surface, causing the stress state at any point on the concrete interface to periodically change over time, thus avoiding excessive stress concentration at fixed nodes due to the standing wave effect.

[0080] The steps S204 to S206 above work together to complete the configuration of the asymmetric excitation array and the establishment of the traveling wave field. The direct technical effect is that the static adsorption force field that was originally uniformly acting on the large-size panel is transformed into a dynamic shear stress field that propagates continuously along the interface, thereby providing an energy source for the micro-adhesion failure described in the subsequent step S207.

[0081] In step S207, the traveling wave, excited by the asymmetric vibrator and propagating within the panel, generates high-frequency reciprocating shear stress in a thin region immediately adjacent to the interface between the metal panel and the concrete. Since the shear strength of the cured concrete is far lower than the interfacial adhesion strength between the steel formwork and the cement paste, the minute shear displacement induced by the traveling wave—typically on the order of micrometers—is sufficient to induce fracture in the weakest microstructure within the bonding layer.

[0082] As vibrational energy is continuously injected and accumulated, the previously sealed vacuum negative pressure area at the interface is gradually destroyed, and external atmosphere enters the original vacuum cavity through microcracks or pore water, causing the pressure difference on both sides of the interface to disappear. This process achieves active release of the template's adsorption stress, rather than passively relying on hydraulic tension to forcibly break it.

[0083] In step S208, regarding the timing control of the exciter array's activation, the system follows a preset spatially ordered start-stop logic. The activation command issued by the central processing unit does not act on all exciters simultaneously, but rather activates the exciters sequentially in a fan-shaped order that spreads from the geometric center of the template to the surrounding edges.

[0084] This inside-out energy release path design aims to provide a low-resistance release channel pointing towards the edge for the high-density elastic energy accumulated in the central area of ​​the template, thereby orderly guiding the internal mechanical stress to the surrounding area of ​​the template and preventing stress concentration or instantaneous impact overload in local structural weak points due to disordered energy release.

[0085] In step S209, during the operation of the exciter, the central processing unit continuously monitors the current feedback signal of the drive motor of each exciter and compares it with the preset load characteristic curve. Once it is detected that the drive current of a certain exciter drops significantly to near the no-load threshold, the system determines that the adsorption force of the interface in that area has been basically released, and issues a corresponding command to reduce the output amplitude of the exciter or remove it from the operation sequence.

[0086] This approach minimizes the system's total power consumption and cumulative vibration fatigue on the template structure while ensuring effective stripping.

[0087] In summary, step S2 constructs a complete logical closed loop from viscous state determination to high-frequency mechanical energy intervention. This stage uses the quantitative load index output from step S1. As input parameters, they are compared with dynamically generated safety thresholds. By comparing data, high-risk operating conditions that require strong intervention can be accurately identified.

[0088] After confirming the risks, the system did not blindly increase the hydraulic pull-out force. Instead, it controlled an asymmetric excitation array to generate traveling waves with phase interference, inducing shear fatigue failure at the concrete interface on a micrometer-scale displacement scale, thereby actively dissipating the adsorbed stress. This process transforms the unavoidable rigid tearing risk in traditional formwork removal processes into a controllable and orderly interface energy release process.

[0089] The next step is S3. After step S2 effectively releases the vacuum adsorption stress at the interface between the formwork and concrete through traveling wave peeling, the system immediately transitions to step S3, where mechanical vibration energy and hydraulic driving force work in synergy. The technical objective of this stage is to utilize the periodic fluctuations of hydraulic tension to actively create tiny macroscopic relative displacement gaps between the formwork panel and the concrete, thereby creating a smooth mechanical path for the subsequent smooth overall retraction of the formwork.

[0090] In step S301, while maintaining the continuous operation of the asymmetric exciter array as in step S2, the central processing unit sends a drive command to the pressure regulating actuator in the hydraulic circuit. Specifically, the pressure regulating actuator is either a high-frequency proportional relief valve or a high-speed switching solenoid valve.

[0091] These valve assemblies have a step response time of less than 10 milliseconds, enabling them to accurately position the valve opening in real time in response to rapidly changing command signals issued by the central processing unit.

[0092] In step S302, the central processing unit generates a control signal for driving the pressure regulating actuator according to a preset pulse control strategy. This control signal can be a pulse width modulation signal with a variable duty cycle, or an analog voltage or current signal with continuously varying amplitude.

[0093] After receiving the control signal, the pressure regulating actuator performs rapid pressurization and depressurization operations on the hydraulic oil in the piston chamber of the retracting cylinder, thereby causing the pressure amplitude acting in the piston chamber to change rapidly between the preset peak pressure and valley pressure.

[0094] In step S303, the direct result of the pressure pulsation is that the pull-out force exerted by the retracting cylinder on the template frame is no longer a constant value, but rather rapidly switches between a preset peak pull-out force and a lower trough pull-out force. This periodic release and release of the pull-out force manifests macroscopically as a small-amplitude but high-frequency "reciprocating sway" of the template panel relative to the concrete interface.

[0095] In step S304, the specific frequency of the pressure pulsation is set within the range of 15Hz to 35Hz. This frequency range is not arbitrarily selected, but is determined by a self-test program executed at the beginning of the system in step S3. The working principle of this self-test program is as follows: the central processing unit controls the exciter to apply excitation to the template reinforcing rib frame in a frequency sweep manner, while using accelerometers installed on the reinforcing ribs to collect the vibration response signal of the frame, and identifies the low-order natural resonant frequency of the template reinforcing rib frame after fast Fourier transform analysis.

[0096] The central processing unit then locks the operating range of the pulsation frequency within a safe frequency band that avoids the low-order inherent resonant frequency, thereby ensuring that the desired mechanical disturbance effect is generated without inducing resonant fatigue damage to the template structure.

[0097] Step S305: The control waveform for pressure pulsation is a sine wave. The reason for choosing a sine wave is its mathematically continuous differentiable characteristic, which enables a smooth transition of pressure amplitude from trough to peak and back again. This smooth transition effectively avoids hydraulic shocks caused by sudden pressure changes, suppresses pipeline vibration and noise, and also reduces the instantaneous rigid impact load on the template frame.

[0098] In step S306, during the pulsating pull-out process, macroscopically visible micro-gaps first form between the template edge and the concrete. As these gaps appear, external air rapidly enters the previously enclosed cavity where residual negative pressure might have existed, completely dissipating the residual adsorption stress that could not be fully eliminated in step S2. Subsequently, the contact state between the template and the concrete rapidly changes from "surface contact" to "low-friction point contact," significantly reducing the interfacial resistance.

[0099] Step S307: To ensure the controllability and safety of the pulsation process, the central processing unit implements closed-loop monitoring of the current feedback value of the high-frequency proportional relief valve or high-speed switching solenoid valve. The system compares the peak-to-peak value of the actual pressure fluctuation with the preset rated pressure amplitude in real time, and strictly constrains the amplitude deviation of the pressure fluctuation within 5% of the rated value by fine-tuning the amplitude or duty cycle of the control signal. When the deviation exceeds the allowable range, the system will immediately correct the control output to prevent damage to hydraulic lines or cylinder seals due to pressure overshoot.

[0100] In summary, step S3, by switching the hydraulic drive mode from constant tension to high-frequency pulsating pulling, further utilizes the fatigue loosening effect of alternating loads to induce initial detachment gaps at the template edge, building upon the disruption of interfacial adsorption stress already achieved in step S2. Through precise vibration damping design of the pulsating frequency via the system self-test program and smooth pressure transition via sinusoidal waveform control, step S3 achieves the final elimination of residual adsorption force with relatively low peak energy consumption while protecting the integrity of the template frame structure.

[0101] Finally, in step S4, after the initial separation gap is formed at the edge of the template through pulsating pulling in step S3, the system enters step S4, which focuses on displacement speed as the core control target. The technical feature of this stage is that the control basis of the demolding process is upgraded from a single pressure threshold judgment to a multi-parameter closed-loop adjustment that integrates displacement, speed and pressure deviation, ensuring that the template exits the working surface uniformly and horizontally under controlled conditions throughout the entire process.

[0102] Step S401: To achieve accurate sensing of the macroscopic motion state of the template, the system arranges multiple sets of high-precision displacement sensors between the four corners and the midpoint of the long side of the template frame and the fixed support base. The displacement sensors are either draw-wire type displacement sensors or laser rangefinders. The sensors are required to have a resolution of no less than 0.01 mm and a data refresh rate that matches the scanning cycle of the programmable logic controller.

[0103] In step S402, the displacement sensor converts the detected current position value of the template into a digital signal or analog electrical signal, which is then transmitted in real time to the counter module or high-speed capture module of the central processing unit via fieldbus or hardwired. The central processing unit reads the current position value at a fixed time period—for example, every 10 milliseconds—and uses a first-order backward differential algorithm to differentiate the displacement data from two consecutive sampling periods, thereby calculating the instantaneous demolding speed of the template in real time. .

[0104] In step S403, the central processing unit runs a closed-loop adjustment algorithm. The core task of this algorithm is to adjust the instantaneous demolding speed. Maintain at the preset standard demolding speed Nearby. Standard demolding speed. The value is preset according to the construction process specifications and usually falls within the range of 2 mm / s to 10 mm / s. The specific value is input by the construction personnel through the human-machine interface or called from the process parameter library during the system initialization phase.

[0105] Step S404, a specific implementation scheme of the closed-loop control algorithm is a fuzzy PID control architecture. Under this architecture, the central processing unit first calculates the deviation between the instantaneous demolding speed and the standard demolding speed. and the rate of change of deviation The definition is as follows:

[0106]

[0107]

[0108] in, This represents the speed deviation, measured in millimeters per second. This represents the rate of change of velocity deviation, expressed in millimeters per second squared. The central processing unit then calculates the... and Each value is multiplied by a preset quantization factor and mapped to a fuzzy subset defined on the fuzzy domain, such as {negative large, negative medium, negative small, zero, positive small, positive medium, positive large}.

[0109] Step S405: The central processing unit processes the fuzzy control rule table pre-stored in the memory. and Rule matching and inference synthesis are performed. The design of the fuzzy control rule table follows these principles: when the speed deviation... When the value is negative and has a large absolute value—meaning the actual speed is much lower than the set value—the control output should be significantly increased to boost energy input; when the speed deviation... If the value is positive and large in absolute terms—meaning the actual speed significantly exceeds the limit—the control output should be rapidly reduced or even the excitation source should be cut off. Take a typical set of rules as an example: If... For negative large and If the value is negative, then the output increment is... For the sake of righteousness, For positive small, It is zero; if zero and If the value is negative, then the output increment will be used. For negative small, Zero For negative; if For the upright and If the value is positive, then the output increment will be greater. For a large negative For negative small, It is positive small.

[0110] Step S406: The result obtained through fuzzy inference is a fuzzy set. The central processing unit uses the centroid method to perform defuzzification operation on the fuzzy set to obtain the three parameters of the PID controller. , , The real-time incremental correction value is obtained. This incremental correction value is then superimposed on a set of baseline PID parameters to obtain the real-time control parameters adapted to the current instantaneous operating conditions. The output of the PID controller... Calculated by the following formula:

[0111]

[0112] in, The controller calculates the adjustment amount of the output, which, after scaling transformation, is mapped to the adjustment command for the amplitude of the excitation force of the asymmetric exciter, the adjustment command for the hydraulic pulsation frequency, and the adjustment command for the target pressure of the proportional relief valve.

[0113] Step S407, another alternative implementation of the closed-loop control algorithm is a model predictive control architecture. In this architecture, the central processing unit maintains a simplified linear time-invariant predictive model of the demodulation process. The state-space expression of this model is:

[0114]

[0115]

[0116] Wherein, the state vector The input vector includes template displacement and instantaneous velocity. Includes excitation force amplitude command and hydraulic pulsation frequency command, output vector This is the predicted demolding speed. Matrix , , The coefficients are determined offline through system identification experiments and stored in non-volatile memory. Within each control cycle, the central processing unit, starting from the current measured state, performs calculations for the future finite time domain. The output trajectory within each step is predicted using a rolling process, and the optimal control sequence for the current time step is solved by minimizing a quadratic cost function. The general form of this cost function is to minimize the tracking error between the predicted velocity and the set velocity, while penalizing changes in the control input. The first control input in the obtained optimal control sequence is applied to the actuator, and the rolling optimization process is repeated in the next cycle.

[0117] Step S408: Regardless of whether a fuzzy PID or model predictive control architecture is used, the control instructions ultimately output by the central processing unit include the following specific actions: when the instantaneous demolding speed... Below the lower limit of the preset speed threshold - for example, below When the adhesion reaches 50% or less, the system determines that there are still unremoved adhesive areas or local obstructions at the interface. At this point, the central processing unit automatically issues a command to increase the excitation force amplitude of the asymmetric vibrator to 120% of the preset reference value, and simultaneously increases the hydraulic pulsation frequency by 5Hz to 8Hz to enhance the energy input to the remaining adhesive interface.

[0118] Step S409, conversely, when the instantaneous demolding speed Exceeding the upper limit of the preset speed threshold—for example, exceeding When the pressure reaches 150%, the system determines that the interface adhesion force has been completely released and the template has entered a state of near-resistance-free free movement. To prevent the template from uncontrollably surging out or impacting due to inertia, the central processing unit immediately issues a command to stop the operation of all asymmetric vibrators, and at the same time switches the high-frequency proportional relief valve in the hydraulic circuit to the stable pressure output mode, so that the drive mode of the retraction cylinder smoothly transitions from pulsating pulling to constant low-pressure stable retraction.

[0119] In step S410, during the execution of step S4, a real-time monitoring mechanism for the uniformity of force on the retraction cylinders operates in parallel with the closed-loop adjustment of the demolding speed. A pressure sensor is embedded in the piston rod head of each retraction cylinder. The central processing unit polls and scans the load pressure values ​​of all cylinders every 100 milliseconds and calculates the arithmetic mean of the load pressures of all cylinders. When the deviation of the load value of a certain cylinder from the average load value exceeds 15% of the average load, the system initiates the off-center load compensation subroutine.

[0120] Step S411, the off-center load compensation subroutine is implemented by adjusting the electromagnetic proportional flow valves independently associated with each retraction cylinder. For cylinders where the detected load pressure is lower than the average value—indicating a relatively large gap and low resistance between the formwork and concrete on that side—the central processor increases the opening of the corresponding proportional flow valve, increasing the oil intake of that cylinder to accelerate its retraction speed. For cylinders where the load pressure is higher than the average value—indicating high resistance on that side—the central processor decreases the opening of the corresponding proportional flow valve or temporarily closes the oil intake of that cylinder, allowing it to wait for other cylinders to catch up. Through this real-time differential compensation strategy, the formwork frame maintains a horizontal posture during demolding, eliminating the risk of in-plane torsional deformation of the frame caused by differences in resistance at various points.

[0121] In summary, step S4 establishes a closed loop by combining the velocity feedback from the displacement sensor with the control commands from the hydraulic-vibration composite actuator, thus enabling the demolding process to transition from "open-loop empirical operation" to "closed-loop adaptive adjustment." This stage uses the instantaneous demolding speed as the core controlled variable, dynamically coordinating the output ratio of excitation energy and hydraulic pulling force through fuzzy PID or model predictive control algorithms to ensure that the mold plate can smoothly exit at a near-preset ideal speed under different viscosity states.

[0122] Throughout the demolding process, the method also involves a real-time monitoring mechanism for the uniformity of force distribution on the retraction cylinders. A pressure sensor is embedded in the piston rod head of each retraction cylinder. The central processing unit scans the load distribution of all cylinders in real time and compares the pressure deviations between cylinders. When the load difference between different cylinders exceeds a preset deviation threshold (e.g., exceeding 15% of the average load), the system activates an off-center load compensation algorithm. By adjusting the independent proportional flow valves associated with each cylinder, the flow input is increased on the side with less force, while the flow is limited on the side with more force, thereby ensuring that the template frame remains horizontally moved backward during demolding and eliminating in-plane distortion of the frame caused by forced pulling on one side.

[0123] During execution, the system utilizes a local storage module and an industrial IoT gateway to continuously record comprehensive parameters, including pressure signals, displacement signals, vibration frequency, pulsation amplitude, ambient temperature, and concrete age. This data is organized into a structured demolding condition package and uploaded to a cloud database. The central processing unit uses a built-in regression analysis algorithm to extract features from the data from multiple demolding processes, automatically optimizing subsequent demolding strategies. This regression analysis algorithm uses the initial demolding load for each demolding operation. The peak value, the time taken to reach the stable shrinkage stage, and the final apparent quality score of the inner wall of the box culvert are used as optimization target variables. Ambient temperature, concrete age, and water-cement ratio are used as input feature variables. Multiple linear regression or support vector regression is used to fit the functional relationship between the input features and the optimal process parameters. For example, when the system detects from historical data that the ambient temperature is higher than 35 degrees Celsius, the optimal stress release triggering time output by the regression model is 10% to 15% earlier than under normal temperature conditions, and the pre-vibration start time is automatically extended by 15% to 20%, thereby achieving feedforward adaptive adjustment of process parameters under subsequent similar conditions.

[0124] Safety protection logic is integrated throughout the entire implementation process. The central processing unit (CPU) is equipped with multiple pressure protection thresholds. When the hydraulic system pressure abnormally surges to the preset safety limit (e.g., 110% of the system's rated pressure) or the displacement sensor detects no displacement change for 5 consecutive seconds, the CPU immediately shuts off the main oil circuit solenoid valve, executes an emergency stop, sends an audible and visual alarm signal to the operating terminal, and records the coordinates of the abnormal point. At this time, the system automatically locks the hydraulic cylinder position and switches to manual intervention mode. After the template has completely exited the working surface and touched the rear limit switch, the system determines that the demolding task is complete, automatically performs a pressure relief operation, and reduces the hydraulic circuit pressure to a safe value.

[0125] Furthermore, the method also exhibits adaptive characteristics during the template installation and adjustment phase. As the template is advanced towards the pouring position, the programmable logic controller (PLC) drives the retraction cylinder to execute fine-tuning commands. Displacement sensors provide real-time feedback on the relative distance between the template and the reference surface. When the distance enters the fine-tuning range within 50 mm, the system automatically switches to a slow-approach mode, reducing the advancement speed to a preset slow value (e.g., 0.5 mm / s). Contact-type limit switches located at the four corners of the template monitor the tightness of the template's fit, ensuring that installation accuracy errors are controlled within a preset range. After installation, the system automatically collects the current displacement coordinates as the zero point and stores them in memory as the reference for the next demolding calculation.

[0126] At the structural design level, the reinforcing ribs of the large-diameter box culvert formwork are made of high-strength alloy steel with a yield strength of not less than 460 MPa. The mounting base of the asymmetric vibrator is fastened to the reinforcing ribs with high-strength bolts, and high-performance damping pads with a thickness of 3 mm to 5 mm are set at the contact surface. This configuration ensures that the high-frequency components of vibration can be transmitted to the panel through the reinforcing ribs, while effectively absorbing and isolating the fatigue stress effect of high-frequency vibration on the welded parts.

[0127] The method features modular expansion capabilities for box culverts of varying diameters. For extra-large diameter box culverts, the system can achieve coordinated control of over 16 sets of retraction cylinders and 32 sets of vibrators by expanding I / O boards and synchronous control modules. Construction personnel only need to input the structural span, wall thickness, and concrete strength grade of the box culvert on the parameter configuration page of the human-machine interface, and the system can retrieve the corresponding pulsation frequency curves and pressure threshold tables from the pre-stored process library, achieving intelligent one-click parameter deployment.

[0128] On the other hand, the adaptive installation and removal adjustment system for the large-diameter box culvert formwork disclosed in this application includes:

[0129] The sensing layer includes a diffused silicon pressure transmitter installed at the high-pressure oil inlet port of the retraction cylinder, displacement sensors arranged between the four corners and the midpoint of the long side of the template frame and the fixed support base, and a pressure sensor embedded in the piston rod head of each retraction cylinder, which are used to collect pressure data, displacement data and load distribution data respectively.

[0130] The decision-making layer includes a central processing unit (CPU) and a programmable logic controller (PLC). The CPU performs filtering on the acquired data, solves for the initial demolding load, and compares the initial demolding load with a set pressure threshold to generate stress release, pulsating retraction, or closed-loop regulation commands. The PLC establishes a data link with the variable frequency drive of the hydraulic pump station through the industrial real-time Ethernet protocol and receives and executes the commands from the CPU.

[0131] The execution layer includes an asymmetric vibrator located at the cross intersection of the reinforcing ribs on the back of the template structure, a high-frequency proportional relief valve or a high-speed switching solenoid valve installed in the hydraulic circuit, and an electromagnetic proportional flow valve independently associated with each retraction cylinder. The asymmetric vibrator is used to excite the traveling wave mode according to the instructions of the decision layer. The high-frequency proportional relief valve or the high-speed switching solenoid valve is used to quickly pressurize and depressurize the hydraulic oil in the piston chamber of the retraction cylinder. The electromagnetic proportional flow valve is used to adjust the oil inlet of the corresponding cylinder during off-center load compensation.

[0132] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present invention. Therefore, the embodiments should be regarded as exemplary and non-limiting in all respects.

[0133] Furthermore, it should be understood that although this specification describes embodiments, not every embodiment contains only one independent technical solution. This narrative style is merely for clarity. Those skilled in the art should consider the specification as a whole, and the technical solutions in each embodiment can also be appropriately combined to form other embodiments that can be understood by those skilled in the art.

Claims

1. A method for adaptive installation, dismantling, and adjustment of formwork for large-diameter box culverts, characterized in that, Includes the following steps: Step S1: Perform a small preload on the retracting cylinder through the hydraulic system, and collect pressure data using a pressure transmitter; The central processing unit filters the collected pressure data and, in conjunction with the pre-stored mechanical efficiency correction coefficient and the total contact area between the formwork panel and the concrete, calculates the initial demolding load. Step S2: Compare the initial demolding load with the set pressure threshold. When it is greater than or equal to the threshold, start the asymmetric vibrator arranged on the reinforcing rib node on the back of the template. By configuring the vibration phase angle deviation between adjacent vibrators, the traveling wave mode is excited on the template panel. Step S3: While maintaining the operation of the vibrator, the hydraulic oil in the piston chamber of the retracting cylinder is rapidly pressurized and depressurized by the pressure regulating actuator, so that the pressure amplitude alternates between the peak value and the valley value, driving the template panel to reciprocate relative to the concrete interface. Step S4: After a gap is formed between the edge of the template and the concrete, the current position value of the template detected by the displacement sensor is obtained and the instantaneous demolding speed is calculated; The central processing unit runs a closed-loop adjustment algorithm to maintain the instantaneous demolding speed near the preset standard demolding speed, and dynamically adjusts the excitation force amplitude and pulsation frequency according to the speed deviation until the template completely exits the working surface.

2. The adaptive installation and dismantling adjustment method for large-diameter box culvert templates according to claim 1, characterized in that, The specific configuration of the asymmetric exciter in step S2 is as follows: each asymmetric exciter integrates a double eccentric block structure for adjusting the excitation force amplitude and a micro servo motor for independently programmed control of the drive phase angle. When the central processing unit generates multi-channel exciter drive pulses, it applies a phase angle deviation of 15 to 45 degrees between two or more spatially adjacent groups of exciters to break the overall resonance trend of the template panel and form a traveling wave mode with non-uniform spatial distribution and fluctuating time series characteristics. At the same time, the activation command issued by the central processing unit activates the exciters in a fan-shaped sequence that spreads from the geometric center of the template to the surrounding edges.

3. The adaptive installation and dismantling adjustment method for large-diameter box culvert templates according to claim 1, characterized in that, In step S3, the pressure regulation actuator is specifically a high-frequency proportional relief valve or a high-speed switching solenoid valve with a step response time of less than 10 milliseconds. The central processing unit generates a pulse width modulation signal with a variable duty cycle or an analog signal with continuously varying amplitude according to a preset pulse control strategy to drive the pressure regulating actuator. The frequency of the high-speed alternating pressure amplitude is determined by the central processing unit through frequency sweeping to identify the low-order natural resonant frequency of the template reinforcing rib frame, and the working range is locked in the 15Hz to 35Hz frequency band that avoids the low-order natural resonant frequency. The waveform of the pressure amplitude change is a sine wave.

4. The adaptive installation and dismantling adjustment method for large-diameter box culvert templates according to claim 1, characterized in that, The closed-loop adjustment algorithm in step S4 is a fuzzy PID control architecture. The central processing unit first calculates the deviation e between the instantaneous demolding speed and the standard demolding speed, as well as the rate of change ec of the deviation. Then, it multiplies the calculated e and ec by a preset quantization factor and maps them to a fuzzy subset on the fuzzy domain. The central processing unit performs rule matching and inference synthesis based on the pre-stored fuzzy control rule table, and uses the centroid method to perform defuzzification operation on the fuzzy set obtained by inference to obtain the real-time incremental correction values ​​of PID controller parameters Kp, Ki, and Kd. Then, it calculates the control quantities used to adjust the amplitude of the excitation force of the asymmetric exciter, the hydraulic pulsation frequency, and the target pressure of the proportional relief valve.

5. The adaptive installation and dismantling adjustment method for large-diameter box culvert templates according to claim 1, characterized in that, The closed-loop adjustment algorithm in step S4 is a model predictive control architecture. The central processing unit internally maintains a linear time-invariant prediction model with template displacement and instantaneous velocity as state vectors and excitation force amplitude instructions and hydraulic pulsation frequency instructions as input vectors; Within each control cycle, the central processing unit (CPU) starts from the current measured state and performs rolling predictions of the output trajectory within a finite time step in the future. It then solves for the optimal control sequence at the current moment by minimizing a quadratic cost function that minimizes the tracking error between the predicted speed and the set speed.

6. The adaptive installation and dismantling adjustment method for large-diameter box culvert templates according to claim 1, characterized in that, During the execution of step S4, a real-time monitoring mechanism for the uniformity of force on the retracting cylinder is also included: the central processing unit polls and scans the load pressure values ​​obtained by the pressure sensor embedded in the piston rod head of each retracting cylinder at a fixed period, and calculates the arithmetic mean of the load pressure of all cylinders. When the load value of a certain cylinder is detected to deviate from the arithmetic mean by more than 15% of the average load, the off-center load compensation subroutine is activated. By adjusting the opening of the electromagnetic proportional flow valve independently associated with each retraction cylinder, the flow input is increased on the side with less force, and the flow is restricted on the side with more force.

7. The adaptive installation and dismantling adjustment method for large-diameter box culvert templates according to claim 1, characterized in that, The method also includes a demolding strategy self-optimization step: organizing all-dimensional parameters, including pressure signal, displacement signal, vibration frequency, pulsation amplitude, ambient temperature, and concrete age, into a structured demolding condition package and uploading it to a cloud database. The regression analysis algorithm built into the central processing unit takes the peak value of the initial demolding load for each demolding operation, the time taken to reach the stable shrinkage stage, and the surface quality score of the inner wall of the box culvert as optimization target variables, and takes the ambient temperature, concrete age, and water-cement ratio as input feature variables. It fits the functional relationship between the input features and the optimal process parameters through multiple linear regression or support vector regression to automatically optimize the subsequent demolding strategy.

8. The adaptive installation and dismantling adjustment method for large-diameter box culvert templates according to claim 1, characterized in that, The method also includes a multi-level safety protection logic: the central processing unit is set with a multi-level pressure protection threshold. When the hydraulic system pressure abnormally surges to 110% of the system's rated pressure or the displacement sensor detects no displacement change within 5 consecutive seconds, the central processing unit immediately cuts off the main oil circuit solenoid valve to execute an emergency stop and sends an audible and visual alarm signal to the operating terminal. At the same time, it automatically locks the hydraulic cylinder position and switches to manual intervention mode. After the template completely exits the working surface and touches the rear limit switch, the system automatically performs a pressure relief operation to reduce the hydraulic circuit pressure to a safe value.

9. The adaptive installation and dismantling adjustment method for large-diameter box culvert templates according to claim 1, characterized in that, The method is also applicable to the template installation and adjustment stage: during the process of advancing the template to the pouring position, the central processing unit drives the retraction cylinder to execute fine-tuning instructions, and the displacement sensor provides real-time feedback on the relative distance between the template and the reference surface; When the relative distance enters the fine-tuning range of 50 mm, the system automatically switches to slow approach mode, reducing the advancing speed to 0.5 mm / s; Contact-type limit switches placed at the four corners of the template monitor the tightness of the template's fit. After installation, the system automatically collects the current displacement coordinates as the zero point and stores them in memory as the reference for the next demolding calculation.

10. A large-diameter box culvert formwork adaptive installation and dismantling adjustment system, used to execute the method according to any one of claims 1 to 9, characterized in that, include: The sensing layer includes a diffused silicon pressure transmitter installed at the high-pressure oil inlet port of the retraction cylinder, displacement sensors arranged between the four corners and the midpoint of the long side of the template frame and the fixed support base, and a pressure sensor embedded in the piston rod head of each retraction cylinder, which are used to collect pressure data, displacement data and load distribution data respectively. The decision-making layer includes a central processing unit (CPU) and a programmable logic controller (PLC). The CPU performs filtering on the acquired data, calculates the initial demolding load, and compares the initial demolding load with a set pressure threshold to generate stress release, pulsating retraction, or closed-loop adjustment commands. The PLC establishes a data link with the variable frequency drive of the hydraulic pump station via an industrial real-time Ethernet protocol and receives and executes the commands from the CPU. The execution layer includes an asymmetric vibrator located at the cross intersection of the reinforcing ribs on the back of the template structure, a high-frequency proportional relief valve or a high-speed switching solenoid valve installed in the hydraulic circuit, and an electromagnetic proportional flow valve independently associated with each retraction cylinder. The asymmetric vibrator is used to excite traveling wave modes according to the instructions of the decision layer. The high-frequency proportional relief valve or the high-speed switching solenoid valve is used to quickly pressurize and depressurize the hydraulic oil in the piston chamber of the retraction cylinder. The electromagnetic proportional flow valve is used to adjust the oil inlet of the corresponding cylinder during off-center load compensation.

Images