Load regulation method of energy-saving hay chopping and rubbing machine based on intelligent control

Abstract

This invention belongs to the field of agricultural machinery control, specifically relating to a load adjustment method for an energy-saving straw chopper and shredder based on intelligent control, aiming to solve the problems of low operating efficiency, poor load matching, and high energy consumption. The method includes: real-time acquisition of operating parameters using sensor components; filtering and normalizing current data to extract load characteristics and identifying operating conditions using a fuzzy neural network model; adjusting the output frequency of the drive device based on the operating condition using a fuzzy PID algorithm; controlling the system to enter sleep mode by monitoring the current value; implementing closed-loop feedback and adaptive parameter adjustment, optimizing control rules using a genetic algorithm, and performing reverse blocking when the current exceeds the limit. This invention, through high-precision logic control and adaptive adjustment, significantly improves the operational stability and load matching of the equipment, effectively reduces system energy consumption, and extends the service life of the equipment.

Description

Technical Field

[0001] This invention belongs to the field of agricultural machinery control, specifically relating to a load adjustment method for an energy-saving straw chopper and shredder based on intelligent control. Background Technology

[0002] As modern animal husbandry transforms towards intensification and intelligentization, straw choppers and shredders, as key equipment for the resource utilization of straw, directly impact feed palatability and production costs through their processing efficiency. Traditional straw chopper and shredder equipment primarily uses high-speed rotating blades to shear and shred crop straw to achieve material refinement and fiberization. With the continuous expansion of processing scale, the requirements for equipment adaptability to materials and operational stability are increasing. How to optimize power configuration and achieve refined control of the production process through automation has become a core demand for improving operational quality in the current feed processing machinery field.

[0003] The load regulation system is a core component ensuring the continuous and stable operation of the straw chopper and shredder. In actual processing scenarios, due to the unevenness of the type, moisture content, and layer thickness of the input materials, the equipment often faces extremely complex load fluctuations during operation. To balance processing efficiency and equipment lifespan, the control system needs to adjust the transmission rate or chopping gap of the feeding mechanism in real time based on the current feedback or torque changes of the main drive motor, so that the equipment load is always maintained near the rated power range, thereby avoiding mechanical failures and minimizing energy consumption per unit output.

[0004] However, most existing load regulation schemes employ linear control logic with fixed parameters, making it difficult to cope with nonlinear disturbances caused by drastic changes in material properties. Due to the significant hysteresis effect in the feedback mechanism, the system cannot make a predictive response to sudden large-flow feeds, easily leading to clogging of the cutting chamber or even motor burnout. Simultaneously, traditional control strategies lack a comprehensive consideration of overall machine energy loss, often failing to effectively reduce operating power under light or no-load conditions, resulting in severe energy waste and mechanical wear. Furthermore, single load monitoring methods cannot accurately identify complex operating states, making it difficult to achieve optimal energy-saving control while ensuring processing quality.

[0005] Therefore, a load adjustment method for an energy-saving straw chopper and shredder based on intelligent control is desired. Summary of the Invention

[0006] This invention provides a load adjustment method for an energy-saving straw chopper and shredder based on intelligent control, which aims to improve equipment operating efficiency and reduce energy consumption through real-time monitoring and adaptive adjustment.

[0007] This invention provides a load adjustment method for an energy-saving straw chopper and shredder based on intelligent control, comprising the following steps:

[0008] Step 1: Collect equipment operating parameters in real time through sensor components; preferably, the sensor components include a current transformer, a Hall sensor, and an ultrasonic sensor; the current transformer has a preset sampling frequency and a predetermined accuracy level; the ultrasonic sensor has a preset detection range and a preset accuracy, used to monitor the real-time changes in the thickness of the material feed.

[0009] Step 2: Filter and normalize the collected real-time current data to extract load characteristics; preferably, construct a multi-dimensional load feature vector by calculating the current change rate and power factor, and input it into a preset fuzzy neural network model to identify the current operating condition; the operating condition includes no-load state, rated state, heavy-load state and overload warning state.

[0010] Step 3: Based on the identified operating conditions, adjust the output frequency of the drive device using a PID control algorithm with integrated fuzzy rules; preferably, the output frequency is within a preset frequency range, and the adjustment process is triggered by the deviation between the real-time current value and the preset reference value.

[0011] Step 4: Monitor in real time whether the device meets the sleep conditions; preferably, if the current value is lower than the no-load judgment threshold and the duration reaches the no-load confirmation time, the control system enters the preset sleep mode.

[0012] Step 5: Implement closed-loop feedback and adaptive parameter adjustment; preferably, the system monitors the load change trend after adjustment in real time, calculates the load fluctuation variance, and uses a genetic algorithm to optimize the membership function of the fuzzy control rule to achieve adaptive adjustment of control parameters according to the material type and moisture content characteristics.

[0013] Preferably, in step 5, if the real-time current value exceeds the blockage determination current threshold and the duration reaches the blockage confirmation time, the blockage handling procedure is triggered; the blockage handling procedure includes controlling the drive device to reverse for a preset reversal time to clear the blockage material, ensuring that the material is completely cleared and re-enters the processing flow.

[0014] According to the technical solution provided by the present invention, compared with the prior art, the present invention has the following beneficial effects:

[0015] (a) Improve the accuracy of working condition identification and load matching

[0016] This invention integrates multi-source sensing data from current transformers, Hall effect sensors, and ultrasonic sensors to construct a multi-dimensional load feature vector containing normalized current values, current change rate, and power factor. It then utilizes a fuzzy neural network model for deep logic computation, achieving accurate identification of no-load, rated, heavy-load, and overload warning states. Compared to traditional methods relying on a single current threshold, this invention effectively distinguishes complex load characteristics caused by changes in material type, moisture content fluctuations, and uneven feed rates, significantly improving the accuracy and robustness of operating condition identification and laying a solid foundation for subsequent refined control.

[0017] (ii) Achieving stable and efficient operation.

[0018] This invention dynamically adjusts the output frequency of the drive device based on the identified operating conditions using a PID control algorithm incorporating fuzzy rules. By real-time correction of the PID parameters according to the current deviation and the rate of change of the deviation, the system can quickly respond to sudden load changes and maintain precise control under steady-state conditions. Under heavy load conditions, the system automatically reduces the feed rate to prevent blockage of the chopping chamber; under light load conditions, the system appropriately increases the operating frequency to ensure processing efficiency. This adaptive adjustment mechanism effectively suppresses mechanical shocks and speed oscillations caused by load fluctuations, significantly improving the stability of equipment operation.

[0019] (iii) Significantly reduce system energy consumption

[0020] This invention achieves energy-saving effects through multiple mechanisms: First, when the equipment is idle for a confirmed duration, it automatically enters a sleep mode, operates at a reduced frequency, or cuts off the main circuit power supply, maintaining only the weak current power supply to the control system, thus completely eliminating energy waste in the ineffective standby state; Second, through precise load matching, the drive motor always operates in the high-efficiency operating range, avoiding power factor drop and energy loss due to overload or light load; Third, it uses a genetic algorithm to evolve and optimize the fuzzy control rules, with minimizing the total energy consumption of the system as one of the optimization objectives, continuously searching for the optimal combination of control parameters to achieve energy consumption optimization throughout the entire life cycle.

[0021] (iv) Extend the service life of equipment

[0022] The load regulation method of this invention effectively reduces the risk of mechanical failure caused by abnormal operating conditions such as overload and material blockage. During the overload warning stage, the system intervenes in advance to prevent the motor from operating under overload for extended periods. When material blockage occurs, the system automatically executes a reverse cleaning procedure, controlling the motor to rotate in the opposite direction to remove the stuck material, protecting the shredder disc, fixed blade, and transmission mechanism from rigid impacts. Furthermore, by monitoring the load fluctuation variance and current energy entropy, the system can predict the material entanglement trend and blade wear degree, taking preventative measures before failure occurs, significantly extending the lifespan of critical equipment components.

[0023] (v) Enhance adaptability to complex material environments

[0024] To address the complex conditions in agricultural production, such as diverse material types, uneven moisture content, and significant differences in geometric dimensions, this invention introduces an adaptive parameter adjustment mechanism based on a genetic algorithm. This mechanism dynamically modifies the membership function of the fuzzy control rules according to real-time load fluctuation characteristics. For silage materials with high toughness and high moisture content, the system automatically increases the response weight to ensure adjustment sensitivity; for dry and brittle straw materials, the system optimizes the smoothness of adjustment to avoid frequent fluctuations. This self-learning and self-optimizing control capability enables the equipment to maintain optimal operating conditions under different material environments.

[0025] (vi) Support intelligent and collaborative operation modes

[0026] In a preferred embodiment of the invention, comprehensive monitoring of material morphology and the chopping process is achieved by integrating multiple sensing methods such as visual recognition units and acoustic emission sensors. A cluster load coordination mechanism based on edge computing enables multiple devices to work collaboratively, dynamically adjusting the operating cycle according to upstream and downstream conditions, effectively preventing material accumulation and improving the overall line's energy efficiency. These intelligent features not only enhance individual machine performance but also provide technical support for the future development of unmanned feed processing workshops.

[0027] In summary, this invention significantly improves the operational stability and load matching of the straw chopper and shredder through high-precision logic control and adaptive adjustment, effectively reduces system energy consumption, and extends equipment lifespan. It has significant application value and promising prospects in the field of intelligent control of agricultural machinery. Attached Figure Description

[0028] Figure 1 This is a schematic diagram of the overall technical solution architecture of the energy-saving chaff cutter and shredder load adjustment method based on intelligent control proposed in this invention;

[0029] Figure 2 This is a schematic diagram of the core principle framework of the fuzzy neural network-based operating condition recognition and PID frequency regulation in this invention;

[0030] Figure 3 This is a logical flowchart of the real-time acquisition of operating parameters and the construction of multi-dimensional load characteristics in this invention;

[0031] Figure 4 This is a flowchart of the logic for adjusting the output frequency of the driving device based on integrated fuzzy rules in this invention.

[0032] Figure 5 This is a schematic diagram of the multi-level interaction relationship and data flow between the sensor components, control system and drive device in this invention;

[0033] Figure 6 This is a flowchart illustrating the logic of the sleep mode switching and automatic material blockage handling program in this invention. Detailed Implementation

[0034] Example 1

[0035] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to specific embodiments.

[0036] Please see Figure 1 In the load adjustment method of an energy-saving straw chopper and shredder based on intelligent control, step 1 involves real-time acquisition of equipment operating parameters using a sensor assembly. The sensor assembly includes a current transformer, a Hall effect sensor, and an ultrasonic sensor. The current transformer has a preset sampling frequency and a predetermined accuracy level, and is installed at the power input of the drive device to monitor the phase current data of the drive motor in real time. The preset sampling frequency is set to 1000 Hz to ensure that it can capture the current distortion signal caused by instantaneous load fluctuations. The predetermined accuracy level of the current transformer is 0.2, converting a large current signal into a millivolt-level voltage signal through a high-sensitivity electromagnetic induction principle, followed by quantization processing by a high-precision analog-to-digital converter. The Hall effect sensor is located at the end of the main shaft of the straw chopper and shredder, generating a pulse signal proportional to the rotational speed by sensing a magnet mounted on the main shaft. The control system calculates the number of pulses per unit time to obtain the precise rotational speed of the main shaft in real time, achieving a speed resolution of 1 rpm. The ultrasonic sensor is installed above the material inlet, has a preset detection range and preset accuracy, and is used to monitor the real-time changes in the material feed thickness. The preset detection range is 10 cm to 100 cm, and the preset accuracy is 1 mm. The ultrasonic sensor emits ultrasonic pulses and receives the echoes reflected from the material surface. Based on the sound wave transmission time difference, it calculates the real-time distance between the sensor and the material surface, and then calculates the thickness of the material accumulation in the feed hopper.

[0037] In step 2, as Figure 2 and Figure 3As shown, the collected real-time current data is filtered and normalized to extract load characteristics. First, a moving average filtering algorithm is used to smooth the original current sequence, removing instantaneous noise caused by power grid fluctuations or electromagnetic interference. The moving average window width is set to 20 sampling points. Then, the filtered current data is normalized and mapped to the [0,1] interval. Specifically, a multi-dimensional load feature vector is constructed by calculating the current change rate and power factor. The current change rate is the derivative of the difference between the current value at the current moment and the current value at the previous moment with respect to time, reflecting the rate of increase or decrease of load. The power factor is calculated by the phase difference between the voltage and current signals and is used to characterize the operating efficiency of the motor. The constructed multi-dimensional load feature vector is input into a preset fuzzy neural network model to identify the current operating condition. The fuzzy neural network model consists of an input layer, a fuzzification layer, a fuzzy rule inference layer, and an output layer. Its specific construction and training process includes:

[0038] First, the network structure is determined: the number of nodes in the input layer equals the dimension of the multidimensional load feature vector. In this embodiment, there are 3 input nodes, corresponding to the normalized current value, current rate of change, and power factor, respectively. Each input in the fuzzification layer corresponds to multiple membership functions. In this embodiment, each input uses 5 Gaussian membership functions, whose center values ​​and initial widths are determined based on historical operating data using a clustering algorithm (such as fuzzy C-means) to ensure coverage of the entire domain of the input variables. The fuzzy rule layer implements fuzzy inference. Each node represents a fuzzy rule, and the number of rules is determined by the combination of nodes in the fuzzification layer. In this embodiment, the number of rules is 5³ = 125, but it can be reduced to about 50 rules in actual use by filtering based on expert experience or rule reduction algorithms. The premise part of the rule is connected to the fuzzification layer, and the conclusion part is connected to the output layer. The output layer has 4 nodes, corresponding to four operating conditions: no-load, rated, heavy-load, and overload warning. The output layer uses a linear activation function, and the output value represents the confidence level of each operating condition.

[0039] Secondly, the network was trained: historical operating data of the straw chopper and shredder under different working conditions were collected, including multi-dimensional load feature vectors and their corresponding working condition labels, forming a training sample set. The training employed an error backpropagation algorithm, using mean squared error as the loss function, and iteratively adjusting the parameters (center and width) of the membership function of the fuzzy layer and the connection weights of the output layer. During training, adaptive learning rate and momentum terms were used to accelerate convergence and avoid getting trapped in local minima. After training, the network parameters were fixed and deployed in the control system.

[0040] In actual operation, the multi-dimensional load feature vector obtained by real-time calculation is input into the fuzzy neural network model. After forward calculation, the confidence level of each working condition at the current moment can be output. The working condition with the highest confidence level is taken as the recognition result, thereby realizing the accurate identification of the operating condition.

[0041] The operating conditions include no-load, rated, heavy-load, and overload warning states. When the current value is below 20% of the motor's rated current and the material thickness detected by the ultrasonic sensor is close to zero, the model outputs a no-load state. When the current value is between 80% and 100% of the rated current and the speed is stable, it is identified as the rated state. When the current value exceeds 110% of the rated current but does not reach the stall current threshold, it is identified as the heavy-load state. When the current value continues to rise and the speed shows a significant downward trend, an overload warning state is output.

[0042] In step 3, please refer to the following: Figure 4 Based on the identified operating conditions, a PID control algorithm integrating fuzzy rules is used to adjust the output frequency of the drive unit. The drive unit is a high-performance vector control frequency converter, and its output frequency adjustment range is limited to a frequency operating range determined based on the motor's rated frequency and mechanical transmission characteristics, typically set to 20 Hz to 60 Hz. The adjustment process is triggered by the deviation between the real-time current value and a preset reference value. The PID control algorithm integrating fuzzy rules dynamically adjusts the proportional, integral, and derivative coefficients of the PID controller through fuzzy logic. When a heavy load condition is detected, the fuzzy controller, based on the current deviation E and the deviation change rate EC, queries a preset fuzzy control table, increases the proportional coefficient to accelerate the response speed, and simultaneously appropriately reduces the output frequency, thereby reducing the instantaneous load on the chopping chamber by decreasing the conveying speed of the feeding mechanism.

[0043] In the specific implementation of step 3, in order to achieve smooth frequency regulation, the system introduces the following load regulation calculation formula:

[0044]

[0045] in, The real-time output frequency of the drive device. This is the reference frequency for the corresponding operating conditions. The preset current reference value, The current value is collected in real time. , , These are the PID parameters that are corrected in real time using fuzzy rules. This formula ensures that the output frequency can be closed-loop compensated according to the current deviation when the load fluctuates, preventing the motor from stopping due to overload.

[0046] Regarding the specific implementation of the PID control algorithm integrating fuzzy rules, this embodiment further discloses the core rule base of the fuzzy controller. The fuzzy controller adopts a two-dimensional input structure, with the input variables being the current deviation E and its rate of change EC. The fuzzy subsets of E and EC are divided into seven levels: negative large (NB), negative medium (NM), negative small (NS), zero (ZO), positive small (PS), positive medium (PM), and positive large (PB), with the universe of discourse normalized to the interval [-3, 3]. The output variables are the correction values ​​of the PID parameters ΔKp, ΔKi, and ΔKd, whose fuzzy subsets are also divided into seven levels, and the universe of discourse is preset according to the characteristics of the frequency converter and motor. The membership function adopts a trigonometric function, which has the characteristics of simple calculation and good real-time performance.

[0047] Some key fuzzy control rules are shown in the table below. These rules cover the adjustment requirements of the straw chopper and shredder under typical operating conditions:

[0048] E EC ΔKp ΔKi ΔKd Control Strategy Description PB PB PB NB PS Severe overload and exacerbation, rapidly enhance proportional action to quickly reduce frequency, suppress integral to prevent overshoot, and fine-tune derivative to smooth response. PB PS PM NM PS Severe overload but with a gradual trend: moderately increase the proportion and reduce the integral to accelerate steady-state recovery. PM ZO PM NS ZO With moderate overload and stability, the ratio is appropriately increased, and the integral is slightly reduced to eliminate steady-state error. PS NB NS PM NS Slight overload but rapid decrease, reduce proportional gain to avoid over-adjustment, and enhance integral gain to eliminate residual bias. ZO NS ZO PS ZO As the value approaches and decreases, the ratio is maintained, and the integral is slightly increased to improve accuracy. NS NS NM PM NS Slight underload and decrease, reduce the ratio, increase the integral to slowly increase the frequency. NB NB NB PB PS The severe underload worsens, the ratio is significantly reduced, and the integral is enhanced to smoothly restore to the rated state. NB PS NM PM ZO The load is severely underloaded but showing signs of recovery. The proportion should be appropriately reduced and the integral increased to promote recovery.

[0049] In the above rules, the fuzzy values ​​of ΔKp, ΔKi, and ΔKd are defuzzified using the centroid method and then added to the initial PID parameters to obtain the real-time control parameters. For example, when the fuzzy neural network model identifies a heavy load state, the current deviation E is usually positive or large. If the feed thickness is still large at this time, causing EC to be positive, the controller queries the first or second rule and outputs a large positive ΔKp and a negative ΔKi, thereby quickly reducing the inverter output frequency, reducing the feed speed, and preventing material blockage. When the current begins to fall (EC becomes negative), the third rule is queried, ΔKp is appropriately reduced, and ΔKi is slightly increased, so that the frequency smoothly recovers and avoids new shocks. When an no-load state is identified, E is negative and large. If EC is also negative and large (current remains low), the seventh rule is queried, significantly reducing the proportional action and enhancing the integral action, so that the frequency slowly rises to the standby frequency and avoids frequent start-stop. Through this set of rules, the system achieves adaptive and refined adjustment for complex material conditions.

[0050] like Figure 5 As shown, a multi-level interactive relationship and data flow are formed between the sensor components, the control system, and the drive device: the raw signals collected by the sensors are preprocessed and then uploaded to the control system. The control system generates decision commands through internal algorithms and finally sends them to the drive device for execution, thus forming a complete closed-loop control link.

[0051] Specifically, Figure 5The core control architecture of this invention is demonstrated. This architecture is divided into three layers from top to bottom: a sensing layer, a control layer, and an execution layer. The sensing layer consists of current transformers, Hall effect sensors, and ultrasonic sensors, responsible for real-time acquisition of raw operating parameters such as motor current, spindle speed, and feed thickness. These parameters are processed by a signal conditioning circuit and then sent to the central processing unit (CPU) of the control layer. The CPU integrates a working condition identification module and a control logic module. The working condition identification module judges the current operating state based on a fuzzy neural network model, while the control logic module calls a fuzzy PID algorithm to generate optimized frequency adjustment commands based on the identification results. These commands are then sent to the drive device in the execution layer, namely a high-performance vector control inverter. The inverter adjusts its output frequency according to the command, thereby controlling the operating speed of the drive motor and the feeding mechanism, achieving precise adjustment of the equipment load. Simultaneously, the inverter's operating status and the motor's actual response are fed back to the CPU in real time via sensors, forming a complete closed-loop control circuit. Through this multi-level interactive design, the system can quickly respond to load changes, ensuring that the straw chopper and shredder always operates at a high-efficiency, energy-saving operating point.

[0052] In step 4, as Figure 6 As shown, the system monitors in real time whether the device meets the sleep conditions. If the current value is lower than the no-load judgment threshold and the duration reaches the no-load confirmation time, the control system enters a preset sleep mode. The no-load judgment threshold is set to 1.2 times the motor's no-load current, and the no-load confirmation time is set to 180 seconds. During the monitoring process, the system starts an incrementing counter, checking the current status once per second. If the current remains below the threshold, the counter increments; once the current exceeds the threshold, the counter is immediately reset to zero. When the counter value reaches 180, the system determines that it is currently in an invalid standby state, and then controls the drive device to reduce its frequency to the lowest operating frequency or directly cuts off the power supply to the main circuit contactor, retaining only the low-voltage power supply to the control system, thereby achieving energy saving.

[0053] In step 5, closed-loop feedback and adaptive parameter adjustment are implemented. The system monitors the load change trend after adjustment in real time and calculates the load fluctuation variance. The load fluctuation variance reflects the uniformity of material feeding. A genetic algorithm is used to optimize the membership function of the fuzzy control rules, achieving adaptive adjustment of control parameters based on material type and moisture content characteristics. Specifically, the genetic algorithm encodes the parameters of the fuzzy membership function into chromosomes, using the minimization of system energy consumption and the maximization of processing efficiency as multi-objective fitness functions. Through evolutionary operations such as selection, crossover, and mutation, it continuously searches for the optimal combination of control parameters in the background. For example, for silage corn stalks with high moisture content, due to their high toughness and high chopping resistance, the genetic algorithm automatically increases the response weight of the fuzzy controller, making frequency adjustment more sensitive.

[0054] In the execution logic of step 5, if the real-time current value exceeds the blockage judgment current threshold and the duration reaches the blockage confirmation time, the blockage handling program is triggered. The blockage judgment current threshold is set to 2.5 times the rated current of the motor, and the blockage confirmation time is set to 2 seconds. Once the triggering condition is met, the system immediately determines that a serious blockage has occurred, and the blockage handling program is activated. The blockage handling program includes controlling the drive device to reverse for a preset reversal time (sufficient to exit the blockage material) to clear the blockage material. The preset reversal time is set to 5 seconds, and the reversal frequency is set to 15 Hz. During the reversal, the feed roller rotates in the opposite direction, pushing the excess material stuck between the shredder and the fixed blade back to the feed hopper. After the reversal is completed, the system controls the drive device to stop for 1 second, and then tries to rotate forward again. If the current returns to the normal range after forward rotation, processing continues; if the current still exceeds the limit, the reversal program is repeated until the material is completely cleared and re-enters the processing flow. If the blockage cannot be eliminated after 3 consecutive reversals, the system will issue an audible and visual alarm signal and stop the machine, awaiting manual intervention.

[0055] To further improve the robustness of the system, this embodiment introduces an energy entropy calculation method in the feature extraction stage, and the calculation formula is as follows:

[0056]

[0057] in, The energy entropy of the current signal. For the current signal at the 1st The probability distribution of energy within a frequency band, with the base of the logarithm (log) typically taken as 2 or e, corresponds to the units of entropy being bits (bit) and nat (nat), respectively. By monitoring changes in energy entropy, the system can more accurately predict the entanglement trend of materials and fine-tune the output frequency in advance before a drastic change in current occurs.

[0058] In a specific application example, when an operator feeds a batch of dry straw into the straw chopper and shredder, the ultrasonic sensor detects a rapid increase in the feed thickness due to the low density and brittle texture of the straw. The current transformer in step 1 collects the current rise signal, and the fuzzy neural network in step 2 identifies that the machine is under heavy load. At this point, the PID algorithm in step 3, based on fuzzy rules, lowers the output frequency of the drive device from 50 Hz to 42 Hz, thus reducing the feed speed and ensuring that the amount of material in the chopping chamber remains within the rated processing capacity. As the material is gradually processed, the current value drops, and the system automatically restores the frequency to 50 Hz. If the material supply is interrupted for more than 180 seconds, the sleep program in step 4 is activated, and the equipment enters a low-power state. If large, hard foreign objects are mixed in during processing, causing the chopping disc to jam, the blockage handling program in step 5 will respond within 2 seconds, reversing the motor to remove the foreign objects and protecting the blades and motor from damage.

[0059] Example 2

[0060] In another embodiment of the load adjustment method for an energy-saving straw chopper and shredder based on intelligent control, the sensor components and control strategies have been further refined for large-scale industrial production scenarios.

[0061] In step 1, in addition to the current transformer, Hall sensor, and ultrasonic sensor, the sensor assembly also includes an infrared moisture sensor and a vibration acceleration sensor. The infrared moisture sensor is installed on the side of the feed conveyor belt, using infrared light of a specific wavelength to irradiate the material and receive the reflected spectrum. Based on the absorption characteristics of infrared energy by moisture, it outputs the percentage moisture content of the material in real time. The vibration acceleration sensor is installed on the bearing housing of the crushing chamber to monitor the intensity of mechanical vibration under high load operation. All sensor data is transmitted to the central processing unit via an isolated RS485 bus, with sampling synchronization error controlled within 10 microseconds.

[0062] In step 2, the collected multi-source sensor data undergoes deep fusion processing. In addition to filtering and normalizing the current data, infrared moisture data and vibration data are incorporated into the load feature vector. The structural fatigue load of the equipment is assessed by calculating the root mean square value and peak factor of the vibration signal. The fuzzy neural network model uses moisture content as an important correction factor when identifying operating conditions. Because high-moisture-content materials generate greater viscous resistance during the kneading process, the model lowers the current threshold for rated conditions, thus intervening in frequency regulation earlier.

[0063] In step 3, the output frequency adjustment of the drive device employs a strategy combining graded pre-control and dynamic fine-tuning. When the ultrasonic sensor detects a step increase in the feed thickness, the system does not wait for current feedback but directly reduces the output frequency of the drive device in advance based on the preset material feedforward compensation model. This combination of feedforward control and feedback PID control effectively suppresses system oscillations caused by sudden load changes. The frequency adjustment accuracy of the drive device is improved to 0.01 Hz, ensuring a smooth processing flow.

[0064] In step 4, the logic for determining the sleep condition incorporates an energy efficiency ratio (EER) evaluation index. The system not only monitors the current value but also calculates the processing output per unit of energy consumption in real time. If the current feed is found to be extremely sparse, causing the EER to fall below a preset threshold, the system will suggest entering an "intermittent operation mode" even if the current does not reach the absolute no-load threshold. In intermittent operation mode, the feeding mechanism accumulates a certain amount of material before processing it centrally, and then immediately goes into sleep mode after processing, thus avoiding energy waste caused by prolonged low-load operation.

[0065] In step 5, the adaptive parameter adjustment process employs a parallel genetic algorithm. The control system utilizes a built-in multi-core processor to run evolutionary calculations for multiple populations in the background. The system monitors the power spectral density of load fluctuations in real time and adjusts the mutation probability of the genetic algorithm based on the spectral distribution characteristics. If the load fluctuations exhibit high-frequency characteristics, it indicates that the material crushing resistance is extremely unstable, and the system increases the mutation probability to quickly find more robust control parameters. Simultaneously, an "oscillation cleaning" step is added to the blockage handling procedure. When a single reverse rotation fails to clear the blockage, the drive unit controls the motor to generate high-frequency, small alternating forward and reverse movements, using inertial impact force to loosen the blockage material. A long-duration reverse rotation is then performed, significantly improving the success rate of automatic blockage clearing under complex operating conditions.

[0066] In a specific application example, this embodiment is used in a large-scale silage processing plant. When processing fresh alfalfa with a moisture content of 65%, the infrared moisture sensor feeds data back to the control system. The fuzzy neural network in step 2 automatically lowers the overload warning threshold. When local accumulation of material at the feed inlet causes a sudden change in the ultrasonic sensor value, the feedforward control in step 3 immediately reduces the inverter frequency by 5 Hz. The ammeter shows that the current fluctuates only slightly before returning to stability, avoiding circuit breaker tripping due to overload. During nighttime operation breaks, due to the discontinuous manual feeding, the intermittent operation mode in step 4 effectively reduces idling energy consumption by approximately 30%.

[0067] Example 3

[0068] In another embodiment of the load adjustment method for an energy-saving straw chopper and shredder based on intelligent control, the system's anti-interference capability and energy management efficiency are optimized, taking into account the mobile agricultural machinery operation environment.

[0069] In step 1, considering the unstable voltage of the mobile power supply, the current transformer is equipped with a dynamic reference compensation circuit. The system monitors grid voltage fluctuations in real time and performs proportional compensation on the collected current values ​​to ensure the accuracy of load characteristic extraction. The Hall sensor adopts a dual-redundant design, with two sets of pulse signals XORed to prevent false speed alarms caused by severe field vibrations. The ultrasonic sensor adds a temperature compensation unit, correcting the sound velocity by measuring the ambient temperature to ensure constant accuracy in material thickness monitoring under different seasons and time periods.

[0070] In step 2, the filtering algorithm employs recursive median filtering to resist impulse interference. This algorithm effectively protects the edges of characteristic signals from blurring, addressing the common short-term mechanical impact noise in agricultural machinery operations. A motor temperature rise model is introduced when constructing the multidimensional load feature vector. The motor winding temperature is estimated by integrating the square of the current over time. When the temperature rise rate is too fast, even if the current is within the rated range, the fuzzy neural network will determine the operating condition as a "thermal overload warning," forcing the system to enter a load reduction operation mode.

[0071] In step 3, when adjusting the drive unit using the PID control algorithm with integrated fuzzy rules, a maximum current limiting step is added. The output frequency adjustment step of the drive unit is limited according to the real-time output capacity of the battery pack or on-board generator. While adjusting the frequency, the system also synchronously adjusts the speed ratio between the feed roller and the shredder disc. Through the synergy of mechanical speed change and electronic speed regulation, optimal shredding effect with low energy consumption is achieved.

[0072] In step 4, the sleep mode is refined into two levels: "standby sleep" and "deep sleep". If the idle time reaches the first preset duration, the system shuts off power output and enters standby sleep; if the duration reaches the second preset duration, the system saves all current operating parameters to non-volatile memory and shuts down the main controller's peripheral interface, entering deep sleep. The wake-up process is triggered by the photoelectric sensor at the feed inlet. Once material is detected, the system quickly starts up and resumes its previous operating state within 500 milliseconds.

[0073] In step 5, adaptive parameter adjustment incorporates cloud-based big data analytics. The equipment uploads local load fluctuation data, material types, and corresponding optimal control parameters to a cloud server via its onboard communication module. The cloud server aggregates operational data from different regions, trains a more general control model using deep learning algorithms, and periodically updates the local control system. Regarding material blockage handling, the system adds a load prediction function. By analyzing the second derivative of current changes, it can preemptively execute short-term shutdowns or deceleration actions within microseconds before a blockage occurs, achieving a shift from "post-event handling" to "pre-event prevention."

[0074] In a specific application example, this embodiment is used for tractor-trailer-driven mobile hay chopping operations. When traversing uneven fields, the dual-channel Hall effect sensors effectively filter out false speed pulses caused by vibration. When encountering materials mixed with hard clods of soil, the temperature rise model in step 2 detects the localized thermal effect caused by a surge in instantaneous current, and the predictive function in step 5 controls the feed roller to stop before the blades are damaged, subsequently discharging the clods through a slight reverse rotation. During long-distance relocation, the system's deep sleep mode reduces battery self-discharge to the microampere level, ensuring the equipment's long-term endurance.

[0075] Example 4

[0076] In another embodiment of the load adjustment method for an energy-saving straw chopper and shredder based on intelligent control, the ability to perceive and respond to the morphological characteristics of materials is enhanced for the refined processing of high-fiber materials.

[0077] In step 1, the sensor assembly further integrates a vision recognition unit. This unit is mounted above the feed hopper and captures image information of the material using a high-definition camera. The control system utilizes image processing algorithms to identify the material's geometry, branching pattern, and roughness. These morphological parameters complement the thickness data acquired by the ultrasonic sensor, forming a comprehensive perception of the feed load. The current transformer employs a Hall effect closed-loop current sensor with an extremely wide frequency response, capable of capturing the weak high-frequency current ripples generated each time the shredding blade strikes the material.

[0078] In step 2, the extraction of load characteristics extends from the time domain to the frequency domain. By performing a Fast Fourier Transform (FFT) on the current signal, characteristic peaks related to the cutting frequency of the cutting tool are extracted. The amplitude variations of these peaks can be used to accurately determine the degree of tool wear. The fuzzy neural network model uses the tool wear factor as a weighting coefficient when identifying operating conditions. When the tool becomes dull, leading to increased cutting resistance, the system automatically lowers the rated load setting to protect the motor from fatigue damage.

[0079] In step 3, the PID control algorithm integrating fuzzy rules introduces a nonlinear compensation stage. Addressing the issue of insufficient torque in the low-frequency range, the system automatically activates the torque boost function based on fuzzy inference results. By compensating the V / F curve of the frequency converter, it ensures sufficient chopping power even under low-speed, heavy-load conditions. The output frequency adjustment references not only the current deviation but also the predicted material density output by the visual recognition unit, achieving a higher-dimensional feedforward adjustment.

[0080] In step 4, the monitoring of sleep conditions is linked to the production planning and management system. If the system receives an instruction that the day's processing tasks have been completed and the current value is below the no-load judgment threshold, it will directly enter sleep mode without delay. During sleep, the system will periodically start a self-test program to check the working status of each sensor and the oil level of the lubrication system to ensure reliability upon the next startup.

[0081] In step 5, the adaptive parameter adjustment employs a hybrid strategy of particle swarm optimization (PSO) and genetic algorithm. The fast convergence of PSO is used for initial online optimization, followed by offline fine-tuning using the global search capability of the genetic algorithm. The system dynamically switches optimization strategies based on the nonlinear characteristics of real-time load fluctuations. In the material blockage handling process, acoustic emission detection technology is introduced. Acoustic emission sensors mounted on the casing capture ultrasonic signals generated by material compression and fracture. When the acoustic emission signal intensity exceeds a preset safety threshold, the system determines that the material density is too high, posing a blockage risk, and immediately triggers preventative reversal or deceleration.

[0082] In a specific application example, when processing high-fiber dry corn stalks, the visual recognition unit identified that the stalks had a large diameter and a hard texture. Step 2, through frequency domain analysis, detected a shift in the cutting characteristic peak, indicating slight wear on the cutting blade. Step 3, based on this, automatically increased the incremental step size of the PID controller and activated a 5% torque boost. During processing, due to the stalks becoming damp and clumping together locally, the acoustic emission sensor detected abnormal high-frequency oscillations. Step 5, before the current reached the current threshold for determining material blockage, preemptively halved the feeding speed, successfully avoiding a potentially serious material blockage incident.

[0083] Example 5

[0084] In another embodiment of the load adjustment method for an energy-saving straw chopper and shredder based on intelligent control, a load coordination mechanism based on edge computing is constructed for multi-machine collaborative operation environments.

[0085] In step 1, the sensor components of a single device connect to the cluster manager via an industrial wireless network (such as ZigBee or LoRa). Current waveform data collected by the current transformer is initially compressed locally, with only the characteristic packets uploaded. Data from the Hall effect sensors and ultrasonic sensors are broadcast at a fixed frequency for reference by neighboring devices. In this way, a single device can not only sense its own load but also obtain real-time information about the material supply status upstream of the production line.

[0086] In step 2, spatial correlation analysis was incorporated into the load feature extraction process. By comparing the current feature vectors of two adjacent straw-cutting and shredding machines, the system can determine whether the current load fluctuation is due to a global fluctuation caused by changes in material properties or a local anomaly caused by a mechanical failure of a single machine. The fuzzy neural network model was upgraded to a distributed structure, with some computational tasks handled by the local controller and complex global optimization logic processed by the edge computing gateway, significantly improving the real-time performance of the identification.

[0087] In step 3, the output frequency adjustment of the drive unit achieves cross-device coordination. When the upstream device decelerates due to heavy load, the downstream device synchronously reduces the output frequency of its feeding mechanism to prevent material accumulation on the conveyor belt. (Reference value for the PID control algorithm) It is no longer a fixed value, but is dynamically allocated based on the total current limit of the entire production line. Through this coordinated adjustment, peak shaving and valley filling of the plant's energy consumption are achieved.

[0088] In step 4, the sleep mode determination logic incorporates a "production line linkage" mechanism. If upstream equipment enters sleep or shutdown mode, downstream equipment, after emptying residual materials, will automatically skip the no-load confirmation time and immediately enter sleep mode. The wake-up logic also supports chain triggering, ensuring the orderly start-up process of the production line and preventing the simultaneous start-up of multiple high-power motors from impacting the power grid.

[0089] In step 5, the adaptive parameter adjustment references a historical processing database of similar materials. The system uses a collaborative filtering algorithm to extract the optimal control template under similar operating conditions from the database. The material blockage handling procedure incorporates a "neighbor machine assistance" strategy. When a machine experiences a material blockage and reverses, its upstream feeding equipment automatically stops conveying and sends a warning to the operator's handheld terminal to prevent further material accumulation. After the reverse clearing is complete, the system automatically calculates the optimal timing for forward rotation to match the overall line's cycle time.

[0090] In a specific application example, in an automated workshop containing five straw-chopping and shredding machines, uneven unloading by the forklift at the feeding end caused a short-term overload on machine No. 1. Through the collaborative mechanism described in Example 5, machines No. 2 through No. 5 fine-tuned their feeding frequencies as machine No. 1 slowed down, smoothly distributing the current fluctuations across the entire production line. When machine No. 3 detected metal foreign objects mixed in the material, triggering an emergency blockage procedure, machines No. 1 and No. 2 immediately stopped, while machines No. 4 and No. 5 continued running until the material inside was emptied. This intelligent load regulation significantly improved the overall operational stability of the system.

[0091] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely illustrative of the principles of the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the present invention as claimed. The scope of protection of this invention is defined by the appended claims and their equivalents.

Claims

1. A load adjustment method for an energy-saving straw chopper and shredder based on intelligent control, characterized in that, Includes the following steps: The equipment's operating parameters are collected in real time by sensor components. The phase current data of the drive motor is obtained by current transformer. The spindle speed is obtained by sensing the pulse signal of the spindle magnet plate by Hall sensor. The material feed thickness is obtained by transmitting and receiving ultrasonic pulses by ultrasonic sensor. The collected real-time current data is filtered by moving average to remove instantaneous noise, and the filtered current data is mapped to a unified interval for normalization. A multi-dimensional load feature vector is constructed by calculating the current change rate and power factor. The multi-dimensional load feature vector is input into a preset fuzzy neural network model. Logical operations are performed through the fuzzification layer, fuzzy rule inference layer and output layer inside the model to identify the operating conditions including no-load state, rated state, heavy-load state and overload warning state. Based on the identified working conditions, the output frequency of the drive device is adjusted using a PID control algorithm with integrated fuzzy rules. The proportional coefficient, integral coefficient, and derivative coefficient are dynamically adjusted by querying a preset fuzzy control table based on the current deviation and the rate of change of deviation. The conveying speed of the feeding mechanism is adjusted by changing the output frequency of the drive device through closed-loop compensation. Real-time monitoring of equipment operating status; when the current value is continuously lower than the no-load judgment threshold and the duration reaches the no-load confirmation time, control the drive device to reduce the frequency or cut off the main circuit power supply to enter the preset sleep mode. The system monitors load change trends in real time and calculates load fluctuation variance. It uses a genetic algorithm to perform evolutionary optimization of the membership function of the fuzzy control rules to achieve adaptive adjustment of control parameters. When the real-time current value exceeds the blockage judgment current threshold and the duration reaches the blockage confirmation time, it executes a blockage handling procedure, including controlling the drive device to reverse for a preset reversal time, retracting the material, and restarting the forward rotation.

2. The load adjustment method for an energy-saving straw chopper and shredder based on intelligent control according to claim 1, characterized in that, The specific process of acquiring the device's operating parameters in real time through sensor components includes: The current transformer is installed at the power input terminal of the drive device to monitor the phase current signal of the drive motor in real time, and the voltage signal generated by electromagnetic induction is quantized into a digital sequence using a high-precision analog-to-digital converter, wherein the sampling process follows a preset sampling frequency and a predetermined accuracy level. The Hall sensor is placed at the end of the main shaft of the straw chopper and shredder. It generates a pulse signal proportional to the rotational speed by sensing the rotational trajectory of the main shaft magnet. The control system obtains the real-time rotational speed of the main shaft by calculating the number of pulses per unit time, and the rotational speed resolution reaches a preset unit accuracy. The ultrasonic sensor is installed above the material inlet. It emits ultrasonic pulses and receives the echoes reflected from the material surface. The real-time distance between the sensor and the material surface is calculated based on the time difference of sound wave propagation in the air. The real-time accumulation thickness of the material in the hopper is calculated by combining the geometric dimensions of the hopper. The ultrasonic monitoring process covers a preset detection range.

3. The load adjustment method for an energy-saving straw chopper and shredder based on intelligent control according to claim 1, characterized in that, The specific process of performing moving average filtering on the acquired real-time current data to remove instantaneous noise includes: The window width of the moving average filtering algorithm is set to a preset number of sampling points to smooth the original current sequence and filter out instantaneous pulse noise caused by power grid fluctuations or electromagnetic interference. Normalization mapping is performed on the filtered current data to linearly transform the current amplitude to a preset numerical range; The process of constructing a multidimensional load feature vector includes: calculating the time derivative of the current value at the current moment with respect to the current value at the previous moment to obtain the rate of change of current, which reflects the instantaneous rate characteristics of load change; By collecting the phase difference between voltage and current signals, the power factor of the drive motor is calculated. The normalized current value, current change rate, and power factor are used as components to construct a load feature vector in a multi-dimensional space, which characterizes the operating performance of the motor.

4. The load adjustment method for an energy-saving straw chopper and shredder based on intelligent control according to claim 1, characterized in that, The specific logic for identifying the operating condition status includes: When the detected current value is lower than the no-load determination threshold of the motor's rated current, and the material feed thickness detected by the ultrasonic sensor is close to zero, the fuzzy neural network model determines that the current state is no-load. When the detected current value is within the rated load range of the motor's rated current, and the fluctuation range of the spindle speed is lower than the speed stability judgment threshold, the current state is determined to be rated. When the detected current value exceeds the overload judgment threshold of the motor's rated current, but does not reach the preset stall current limit, the current state is determined to be overloaded. When the detected current value shows a continuous upward trend, and the spindle speed shows a synchronous downward trend of a preset magnitude, the fuzzy neural network model outputs an overload warning status.

5. The load adjustment method for an energy-saving straw chopper and shredder based on intelligent control according to claim 1, characterized in that, The specific process of adjusting the output frequency of the drive device using the PID control algorithm with integrated fuzzy rules includes: The drive device is controlled to operate in vector control mode, and the adjustment range of its output frequency is limited to the frequency operating range determined based on the rated frequency of the motor and the mechanical transmission characteristics. The deviation E between the current value and the preset reference value and the rate of change EC of the deviation are calculated in real time. The deviation E and the rate of change EC are input into the fuzzy controller and mapped to the preset universe of discourse through fuzzification processing. Query the preset fuzzy control rule table and dynamically output the proportional coefficient correction value, integral coefficient correction value and derivative coefficient correction value of the PID controller; The real-time target output frequency of the drive unit is calculated based on the corrected PID parameters. When a heavy load condition is detected, the material conveying speed of the feeding mechanism is reduced by increasing the proportional coefficient response weight and decreasing the output frequency, thereby reducing the instantaneous load pressure in the chopping chamber.

6. The load adjustment method for an energy-saving straw chopper and shredder based on intelligent control according to claim 1, characterized in that, The specific process of monitoring the operating status of the real-time monitoring device and entering a preset sleep mode includes: An incrementing counter is started internally by the system to periodically check the current value status at a preset monitoring cycle; If the current value is lower than the no-load determination threshold, the counter performs an accumulation operation; If the current value exceeds the no-load determination threshold, the counter will immediately be cleared. When the counter value accumulates to the preset value corresponding to the no-load confirmation time, the system determines that the equipment is in an invalid standby state. Subsequently, it controls the drive device to reduce the frequency to the preset minimum operating frequency and cuts off the power supply to the main circuit contactor, maintaining only the power supply to the weak current part of the control system.

7. The load adjustment method for an energy-saving straw chopper and shredder based on intelligent control according to claim 1, characterized in that, The specific process of using a genetic algorithm to perform evolutionary optimization of the membership function of fuzzy control rules includes: The center value and width parameter of the membership function in the fuzzy control rule are encoded into the chromosome sequence of the genetic algorithm; a composite fitness function is constructed with the objectives of minimizing the total energy consumption of the system and maximizing the material processing efficiency to evaluate the merits of the current combination of control parameters. During background operation, selection, crossover, and mutation evolution operations are performed to search for the optimal membership function configuration within the parameter space; For materials with different moisture content and toughness characteristics, the genetic algorithm is adaptively adjusted by monitoring the distribution characteristics of load fluctuation variance to dynamically correct the fuzzy control parameters and improve the system's operational stability in complex material environments.

8. The load adjustment method for an energy-saving straw chopper and shredder based on intelligent control according to claim 1, characterized in that, The specific sequence of actions for executing the material blockage handling procedure includes: When the real-time current value exceeds the blockage judgment current threshold and the duration reaches the blockage confirmation time, the system determines that a mechanical blockage fault has occurred and immediately blocks the positive pulse output of the drive device. The control drive device is switched to reverse mode, and the motor is driven to rotate in the opposite direction according to the preset reverse frequency and preset reverse duration, which drives the feed roller to return the material stuck between the shredding disc and the fixed blade to the feed hopper area. After the reversal is completed, the control drive device stops running for a preset braking time, and then attempts to rotate forward again at a preset starting frequency; If the current value returns to the rated current range after forward rotation, the load adjustment logic continues to be executed; if the current value still exceeds the limit, the reverse cleaning action is repeated until the material is completely removed or the preset maximum number of attempts is reached.

9. The load adjustment method for an energy-saving straw chopper and shredder based on intelligent control according to claim 1, characterized in that, The method also includes a load forecasting and condition assessment process based on multi-source sensing: The visual recognition unit installed above the feed hopper captures the image features of the material, and the geometric size and branch roughness of the material are identified by the image processing algorithm. The recognition result is then used as a feedforward compensation signal input to the control system. By performing a fast Fourier transform on the real-time current signal, characteristic peaks related to the cutting frequency of the shredding blade are extracted. The amplitude evolution trend of the characteristic peaks is monitored to assess the wear degree of the blade, and the judgment criteria for the rated load are automatically adjusted according to the wear degree. Acoustic emission sensors are used to capture acoustic emission signals generated during the material chopping process. By analyzing the energy distribution characteristics of the signals, the entanglement trend of the material can be predicted, and the output frequency of the drive device can be finely adjusted in advance before the current changes abruptly.

10. The load adjustment method for an energy-saving straw chopper and shredder based on intelligent control according to claim 1, characterized in that, The method also includes a cluster load coordination process based on edge computing: The load feature vector of a single device is uploaded to the edge computing gateway through the industrial wireless network, and a multi-machine collaborative load distribution model is established in the cluster manager. When upstream equipment on the production line slows down due to heavy load conditions, the edge computing gateway synchronously sends frequency adjustment commands to downstream equipment on the production line to coordinate the operating rhythm of each feeding mechanism and prevent material from accumulating in the conveying process. The system calculates the energy efficiency ratio of the entire production line in real time and dynamically switches the operating status of a single piece of equipment between rated operation mode and intermittent operation mode based on the continuity of material supply. When executing the material blockage handling procedure, the neighboring machine assistance strategy is triggered through the edge computing gateway to automatically stop the material conveying of the upstream equipment and send a warning signal to the remote monitoring terminal.

Images