Watermelon ripeness nondestructive detection method and system based on water immersion density method

Abstract

This invention discloses a non-destructive testing method and system for watermelon maturity based on the immersion density method, specifically relating to the field of precision physical quantity measurement and non-destructive testing technology. It addresses the problem in existing automated immersion measurements where inaccurate liquid level data due to dynamic process interference affects density calculation accuracy. The method first acquires the basic weight and liquid level data of the watermelon under both unloaded and loaded conditions. Then, it controls the immersion of the watermelon while simultaneously acquiring the driving signal and liquid level change sequence. Next, it identifies the dynamic stages of the immersion process and uses recursive graph analysis to determine the consistency of the driving and liquid level response dynamics. For the data subset approaching equilibrium, it performs multi-scale complexity and long-range correlation analysis to intelligently determine whether the liquid level has reached a physically stable state. Finally, only under the condition of stability is the watermelon density accurately calculated using stable liquid level data combined with weight data. This effectively filters out dynamic interference from immersion, significantly improving the accuracy of density measurement.

Description

Technical Field

[0001] This invention relates to the field of precision measurement and non-destructive testing of physical quantities, and in particular to a non-destructive testing method and system for watermelon maturity based on the water immersion density method. Background Technology

[0002] In the quality sorting and maturity testing of watermelons, the application of non-destructive testing technology is crucial. The water immersion density method, based on Archimedes' principle, measures density by measuring the buoyancy of an object in a liquid or the volume of liquid displaced. This is a well-known physical measurement method. When applied to watermelon testing, it typically involves completely immersing the watermelon in water. A distance sensor, such as a level sensor, measures the change in liquid level to calculate the displaced volume. This, combined with the mass previously measured by a mechanical sensor, such as a weighing sensor, is used to calculate the watermelon's density. The maturity is then assessed based on the correlation between density and internal quality. To automate the process, existing solutions generally use mechanical devices to perform the immersion operation, and after immersion, a level sensor collects the liquid level data.

[0003] However, in automated implementation, existing methods typically assume that once the watermelon is submerged and stationary by the mechanical device, the measured liquid level height accurately corresponds to its actual drainage volume. In reality, due to the irregular shape of the watermelon and the dynamic nature of the submersion process, the time from the watermelon's initial contact with the water surface to its complete submersion, the dissipation of internal gases, and the return of the water to calm is a continuous dynamic process. The conventional method of measuring immediately after submersion in existing technologies easily captures unstable interference components from this dynamic process, resulting in random errors in the measured volume value that cannot be eliminated through simple calibration. This severely limits the required accuracy and reliability in real-world continuous and rapid detection scenarios. Summary of the Invention

[0004] This invention addresses the technical problems existing in the prior art by providing a non-destructive testing method and system for watermelon maturity based on the water immersion density method.

[0005] The technical solution of the present invention to solve the above-mentioned technical problems is as follows:

[0006] A non-destructive method for watermelon maturity testing based on the water immersion density method includes:

[0007] S1. Obtain the first liquid level and first weight of the bearing device when it is unloaded and submerged in water, and the second weight of the bearing device after it is loaded with watermelon;

[0008] S2. Control the carrier device to immerse the watermelon in water, synchronously record the drive control signal and collect the liquid level height to form a sequence of liquid level height changes;

[0009] S3. Based on the sequence of changes in liquid level, identify the dynamic stages in the process of immersing watermelons;

[0010] S4. Based on the recursive graph analysis method, determine whether the dynamic similarity between the driving control signal and the liquid level change sequence meets the requirements;

[0011] S5. If the requirements are met, extract the liquid level height data subset corresponding to the target stage used for stability determination in the dynamic stage, analyze the correlation pattern between the complexity and long-range correlation of the liquid level height data subset at multiple scales, and determine whether the liquid level is in a stable state that can be used for density calculation.

[0012] S6. If the watermelon is in a stable state, determine the second water level based on the subset of water level data, and calculate the watermelon density by combining the first water level, the first weight, and the second weight.

[0013] Furthermore, S1 includes:

[0014] When the load-bearing device is in an unloaded state, the load-bearing device is controlled to be completely submerged in water. The first liquid level height is measured by a liquid level sensor, and the first weight of the unloaded load-bearing device is measured by a weighing sensor.

[0015] Next, the watermelon was loaded onto the support device, and the combined weight of the support device and the watermelon was measured using the same weighing sensor.

[0016] Furthermore, S2 includes:

[0017] The control device carries the watermelon into the water to perform the immersion operation. During this process, the drive control signal that drives the movement of the device is recorded synchronously and continuously, and the liquid level height is collected at a fixed sampling frequency by the liquid level sensor.

[0018] From the moment the watermelon begins to touch the water surface, data is continuously recorded and collected until the immersion process is completed and the predetermined collection time is reached. The collected water level data, arranged in chronological order, forms a sequence of water level changes.

[0019] Furthermore, S3 includes:

[0020] Differential calculations are performed on the sequence of liquid level changes to obtain the corresponding sequence of liquid level change rates;

[0021] Based on the numerical distribution and zero-crossing characteristics of the liquid level height change rate sequence, the watermelon immersion process is divided into a dynamic stage that includes at least a rapid change stage and a tendency to equilibrium stage.

[0022] Furthermore, S4 includes:

[0023] Phase space reconstruction was performed on the driving control signal and the liquid level change sequence to obtain the phase space trajectories corresponding to the two time series respectively.

[0024] Each recursion matrix is ​​calculated based on the phase space trajectory. The elements in the recursion matrix indicate whether the phase space state point at the corresponding time is within the set neighborhood radius.

[0025] Recursive quantitative analysis is performed on the two recursive matrices to extract parameters characterizing the dynamic properties of the system.

[0026] Calculate the similarity metric between two parameter sets, compare the similarity metric with a preset threshold, and determine whether the dynamic similarity meets the requirements based on the comparison result.

[0027] Furthermore, the calculation of the respective recursion matrix based on the phase space trajectory includes: for each phase space state point in the phase space trajectory, calculating the Euclidean distance between the corresponding phase space state point and all other phase space state points in the phase space trajectory, and comparing the calculated Euclidean distance with a pre-set neighborhood radius; if a certain Euclidean distance is less than or equal to the corresponding neighborhood radius, then the corresponding position in the recursion matrix is ​​marked as the first preset value, indicating that recursion has occurred between the corresponding two phase space state points; if it is greater than the corresponding neighborhood radius, then it is marked as the second preset value, indicating that no recursion has occurred; in this way, all phase space state points are traversed to form a complete recursion matrix.

[0028] Furthermore, S5 includes:

[0029] From the identified dynamic stages, the equilibrium stage is selected as the target stage for stability determination, and the corresponding liquid level height data subset is extracted for the target stage.

[0030] The multi-scale sample entropy of liquid level height data subsets at different time scales is calculated to characterize the complexity at multiple scales. At the same time, the detrended fluctuation analysis scaling index of the corresponding liquid level height data subset at the corresponding scale is calculated to characterize the long-range correlation.

[0031] The covariance between the multi-scale sample entropy sequence and the detrended fluctuation analysis scaling exponential sequence is analyzed as a correlation pattern.

[0032] The correlation pattern is compared with a preset stability criterion to determine whether the liquid surface is in a stable state that can be used for density calculation.

[0033] Furthermore, calculating the detrended volatility analysis scaling index of the corresponding liquid level height data subset at the corresponding scale to characterize long-range correlation includes: cumulatively summing the liquid level height data subset and removing the mean to generate a new profile sequence; dividing the profile sequence into multiple non-overlapping intervals of equal length, fitting the local trend within each interval using the least squares method, and subtracting the local trend from the profile sequence to obtain the detrended subsequence; calculating the average variance of the detrended subsequences within all intervals, and then taking the square root to obtain the volatility function value at the corresponding scale; changing the interval length to correspond to different time scales to obtain volatility function value sequences at different scales; and linearly fitting the volatility function value sequence to the corresponding scale in a double logarithmic coordinate system, with the slope of the fitted line being the detrended volatility analysis scaling index.

[0034] Furthermore, S6 includes:

[0035] Statistical analysis was performed on a subset of liquid level data, and the average liquid level was calculated as the second liquid level.

[0036] The change in water level caused by the immersion of the watermelon is calculated based on the difference between the first and second water level heights, and the volume of water displaced by the watermelon is calculated based on the known cross-sectional area of ​​the container.

[0037] Calculate the weight of the watermelon based on the first and second weights;

[0038] Based on Archimedes' principle, the density of a watermelon can be calculated using its weight and the volume of water it displaces.

[0039] On the other hand, the present invention provides a non-destructive testing system for watermelon maturity based on the water immersion density method, comprising:

[0040] The data acquisition module is used to acquire the first liquid level height and first weight of the carrier when it is unloaded and submerged in water, and the second weight of the carrier after it is loaded with watermelon.

[0041] The sequence formation module is used to control the carrier device to immerse the watermelon in water, synchronously record the drive control signal and collect the liquid level height to form a sequence of liquid level height changes;

[0042] The phase identification module is used to identify the dynamic phases in the watermelon immersion process based on the sequence of liquid level changes;

[0043] The requirement is to have a judgment module that, based on a recursive graph analysis method, determines whether the dynamic similarity between the driving control signal and the sequence of liquid level changes meets the requirements.

[0044] The state judgment module is used to extract the liquid level height data subset corresponding to the target stage for stability judgment in the dynamic stage if the requirements are met, analyze the correlation pattern between the complexity and long-range correlation of the liquid level height data subset at multiple scales, and determine whether the liquid level is in a stable state that can be used for density calculation.

[0045] The density calculation module is used to determine the second liquid level height based on a subset of liquid level height data if the liquid level is in a stable state, and to calculate the watermelon density by combining the first liquid level height, the first weight, and the second weight.

[0046] The beneficial effects of this invention are:

[0047] 1. By introducing detailed analysis and intelligent stability assessment of the immersion dynamic process, the accuracy and reliability of density measurement results are fundamentally improved, thus laying an accurate data foundation for subsequent density-based maturity assessment. Specifically, by synchronously recording the drive control signal and liquid level changes, and using recursive graph analysis technology to verify the dynamic similarity between the two, the system can effectively identify abnormal immersion processes caused by mechanical execution anomalies or external disturbances. This ensures that only experimental data in which the drive and liquid level responses are dynamically coordinated can enter the subsequent analysis process, filtering out the risk of gross errors introduced by actuator step loss, jamming, or external impacts from the source. This enhances the robustness of the entire detection system to accidental disturbances during operation and ensures the consistency of measurement conditions.

[0048] 2. Instead of simply collecting the liquid level height after the immersion process, the system identifies the dynamic stages during immersion based on sequence analysis. Specifically targeting the equilibrium stage data, it introduces multi-scale complexity and long-range correlation pattern analysis to determine whether the liquid level has reached a physically stable state. This allows the system to intelligently capture and wait for the true equilibrium moment when the liquid level has fully relaxed from mechanical disturbances and the influence of internal gas has been largely eliminated. This extracts the most representative subset of stable liquid level height data for the actual watermelon drainage volume, effectively removing inertial or random interference components such as incompletely decaying liquid level fluctuations and the escape of tiny bubbles during the dynamic process. This solves the measurement noise problem inherent in the immersion of irregularly shaped objects, which is difficult to eliminate through fixed delays or simple filtering. Ultimately, this achieves a more accurate and stable measurement of the drained water volume. Attached Figure Description

[0049] Figure 1 This is a flowchart of the non-destructive testing method for watermelon maturity based on the immersion density method of the present invention;

[0050] Figure 2 This is a schematic diagram of the structure of the watermelon maturity non-destructive testing system based on the water immersion density method of the present invention. Detailed Implementation

[0051] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0052] Example 1: Figure 1 This invention presents a non-destructive testing method for watermelon maturity based on the water immersion density method, comprising:

[0053] S1. Obtain the first liquid level and first weight of the bearing device when it is unloaded and submerged in water, and the second weight of the bearing device after it is loaded with watermelon;

[0054] S2. Control the carrier device to immerse the watermelon in water, synchronously record the drive control signal and collect the liquid level height to form a sequence of liquid level height changes;

[0055] S3. Based on the sequence of changes in liquid level, identify the dynamic stages in the process of immersing watermelons;

[0056] S4. Based on the recursive graph analysis method, determine whether the dynamic similarity between the driving control signal and the liquid level change sequence meets the requirements;

[0057] S5. If the requirements are met, extract the liquid level height data subset corresponding to the target stage used for stability determination in the dynamic stage, analyze the correlation pattern between the complexity and long-range correlation of the liquid level height data subset at multiple scales, and determine whether the liquid level is in a stable state that can be used for density calculation.

[0058] S6. If the watermelon is in a stable state, determine the second water level based on the subset of water level data, and calculate the watermelon density by combining the first water level, the first weight, and the second weight.

[0059] S1. Obtain the first liquid level and first weight of the supporting device when it is unloaded and submerged in water, and the second weight of the supporting device after loading watermelons. The specific implementation is as follows:

[0060] Before the formal testing begins, the necessary basic data for calculation needs to be obtained. First, the unloaded support device is measured. This device, driven by a mechanical mechanism, can be raised and lowered; it can take the form of a net or a bracket for holding the watermelon. When measuring the initial liquid level, the unloaded support device is controlled to descend slowly at a low speed, for example, less than 10 centimeters per second, until it is completely submerged in the water-filled container. Complete submersion is determined when the highest point of the support device is at least one set depth below the initial static liquid level, for example, 5 centimeters, and remains stationary at that position for a set time, for example, 3 seconds, until the liquid level stabilizes. At this point, the liquid level data is obtained through a hydrostatic level sensor fixedly installed on the side wall of the container. This sensor operates based on the principle that liquid pressure is proportional to depth; its output electrical signal is converted into a digital signal by an analog-to-digital converter and read by the controller. The controller uses a pre-stored calibration relationship for the sensor to convert the read digital value into a specific liquid level height in millimeters; this value is stored as the initial liquid level height.

[0061] Subsequently, the support device was removed from the water and its surface moisture was drained, and its unloaded weight was measured. Weight measurement was performed using a strain gauge load cell directly mounted on the force path of the support device. The strain gauge inside the load cell deforms under load, causing a Wheatstone bridge to output a voltage signal proportional to the load. This signal is amplified and converted from analog to digital before being read by the controller. The controller uses parameters obtained beforehand through calibration experiments, such as the linear coefficient fitted using standard weights, to convert the reading into a weight value in grams. This value is stored as the initial weight of the support device and correlated with the previously obtained initial liquid level data.

[0062] After completing the no-load parameter measurements, the weight of the watermelon after loading is measured. The watermelon is placed stably in the center of the supporting device and brought to a stable stop. Using the same weighing sensor as described above, under the same signal acquisition and processing conditions, the total weight of the supporting device and the watermelon is measured again and obtained. This weight is defined as the second weight and stored. Although the net weight of the watermelon can be obtained by subtracting the first weight from the second weight, this calculation is not necessary for this step. The first liquid level height, the first weight, and the second weight obtained above are all assigned a unified detection label to ensure correct retrieval in subsequent steps.

[0063] S2. Control the carrier device to immerse the watermelon in water, simultaneously record the drive control signal and collect the liquid level height to form a sequence of liquid level height changes. The specific implementation is as follows:

[0064] First, the control device, carrying the watermelon already placed on it, moves from its initial position into the water container to perform an immersion operation. This movement is executed by a drive mechanism such as a servo motor or stepper motor, and controlled by a controller sending movement commands to the motor driver. These movement commands are a specific form of the drive control signal that needs to be recorded. The drive control signal may include the pulse width modulation duty cycle value, pulse frequency, or real-time position setpoint of the motor controller. For synchronous recording, a high-precision timer is started at the same moment the controller issues the movement command, and data is collected at a fixed sampling frequency. The sampling frequency setting needs to balance signal bandwidth and data volume; for example, 100 Hz can be selected. The system reads and stores the instantaneous values ​​of the movement commands at this frequency, forming a drive control signal sequence that strictly corresponds to time.

[0065] While the carrier moves the watermelon into the water, the liquid level in the container is collected at the same fixed sampling frequency using the same hydrostatic level sensor used in step S1. The analog signal from the level sensor is synchronously acquired via an independent analog-to-digital conversion channel, ensuring that the timestamp of each sampling point is aligned with the sampling timestamp of the drive control signal from the same system clock source to achieve precise time synchronization. The acquisition process is triggered by a timer interrupt service routine in the controller to ensure the accuracy of the sampling interval, thereby obtaining discrete but equally timed instantaneous values ​​of the liquid level.

[0066] The recording and acquisition of the immersion process has clearly defined start and end conditions. The start time is defined as the instant the watermelon begins to contact the water surface. This determination is based on the analysis of the real-time collected liquid level height values. Specifically, when the carrying device lowers the watermelon, before the watermelon touches the water, the liquid level remains at the first liquid level height measured in step S1, and its collected value fluctuates within the sensor noise range. The controller calculates the absolute value of the difference between the newly collected liquid level height value and the first liquid level height in real time. This absolute value of the difference is compared with a pre-set water contact sensitivity threshold. The water contact sensitivity threshold is set based on the background noise amplitude of the liquid level sensor under static water and the maximum fluctuation amplitude that may be caused by minor environmental vibrations. Its specific value is determined through preliminary experiments, for example, by repeatedly measuring the fluctuation range of the liquid level height when stationary and multiplying its upper limit by a safety factor, the water contact sensitivity threshold can ultimately be set to 0.5 mm. When the controller detects that the absolute value of the liquid level height difference at multiple consecutive sampling points is greater than the water contact sensitivity threshold, it determines that the watermelon has begun to contact the water surface. The number of consecutive sampling points here is another criterion used to avoid false triggering caused by a single abnormal sampling point. This number is set to, for example, 3 points. The system marks the timestamp corresponding to the first sampling point in the first batch of sampling points that meet the above conditions as the start time of the immersion record.

[0067] From the start moment, the system continuously and synchronously records the drive control signals and collects the liquid level. The carrier continues to move along the preset trajectory until the watermelon is completely submerged and the movement stops. The termination of data acquisition is controlled by a preset total acquisition duration. This total acquisition duration is set with the start moment as the timing point, and its setting must ensure complete recording of the entire process from the watermelon touching the water, rapid submersion, to the decay of liquid surface sloshing, and finally to physical equilibrium. The specific value of the total acquisition duration is determined based on the observation and statistics of the time required for the liquid level to stabilize after the watermelon is submerged. For example, by analyzing the time distribution required for the rate of change of liquid level to decay to near zero in multiple experiments, and taking its statistical upper limit, the total acquisition duration can be set to 10 seconds. When the high-precision timer reaches the predetermined end of the total acquisition duration, the system stops recording and acquiring all signals.

[0068] Finally, all instantaneous liquid level height values ​​collected from the start time to the end of the total collection period are arranged strictly according to their timestamps, forming an ordered data set, which is defined as the liquid level height change sequence. Each data point in this sequence contains two basic elements: the precise collection time and the corresponding liquid level height value. The drive control signal sequence is also arranged and stored according to the same time base.

[0069] S3. Based on the sequence of changes in liquid level, identify the dynamic stages during the watermelon immersion process. The specific implementation is as follows:

[0070] After obtaining the liquid level change sequence generated in step S2, the dynamic stages of the watermelon immersion process are identified. First, differential calculations are performed on the liquid level change sequence to obtain the liquid level change rate sequence. Since the liquid level change sequence is discrete time-series data, the differential calculation is implemented using the first-order difference method. Specifically, assume the liquid level change sequence contains N height values ​​arranged in chronological order.

[0071] Calculate the rate of change of liquid level height: R(n) = (H(n+1) - H(n)) / ΔT; where R(n) is the calculated rate of change of liquid level height for the nth time, in millimeters per second; H(n) represents the nth height value in the liquid level height change sequence, in millimeters; H(n+1) represents the (n+1)th adjacent height value in the sequence; ΔT represents the sampling time interval between two adjacent height values, in seconds. This time interval is the reciprocal of the sampling frequency; for example, when the sampling frequency is 100 Hz, ΔT equals 0.01 seconds. Perform this calculation on all positions from the first point to the (N-1)th point in the sequence to generate a new data sequence containing N-1 elements, which is the liquid level height change rate sequence. To ensure that the calculated sequence reflects real physical changes rather than noise, the liquid level change sequence can be preprocessed before calculation. For example, a moving average filter can be used to smooth the original data. The width of the filter window is set according to the noise frequency characteristics. For example, a window covering 5 consecutive sampling points can be selected for averaging calculation.

[0072] After obtaining the liquid level change rate sequence, the entire immersion process needs to be divided into different dynamic stages based on its numerical distribution and zero-crossing characteristics. First, the numerical distribution of the liquid level change rate sequence is analyzed. During the entire immersion period, the liquid level change rate will evolve from a large positive value to near zero. A key criterion is to set a stability threshold for the change rate, used to identify when the liquid level change enters a slow equilibrium state. The setting of this stability threshold needs to consider the resolution of the liquid level sensor in a static state and the inherent noise level of the system. The specific acquisition method includes: under the stable conditions of step S1, collecting liquid level data for a period of time and calculating its standard deviation to characterize the system's background noise intensity; then, dividing this standard deviation by the sampling time interval ΔT yields a benchmark value characterizing the equivalent change rate of the background noise; finally, multiplying this benchmark value by a safety factor, such as 2 or 3, the resulting product is set as the stability threshold. The threshold set by this method has clear physical meaning and statistical basis.

[0073] The specific logic for the dynamic phase division is as follows. The rapid change phase begins at the start time of the immersion recording defined in step S2. Its characteristic is that the rate of change of liquid level R(n) is large and consistently positive, indicating that the liquid level is rising rapidly. The termination point of this phase is determined by comparing the zero-crossing feature of the liquid level change rate sequence with the rate of change stability judgment threshold. The zero-crossing point refers to the moment when the rate of change of liquid level R(n) first changes from a positive value to a zero or negative value. To enhance the robustness of the judgment and avoid misjudgments due to instantaneous noise or small fluctuations, the system requires that after the zero-crossing point occurs, the absolute value of the rate of change of liquid level R(n) must remain below the aforementioned rate of change stability judgment threshold for a period of time. For example, after the system detects that R(n) is less than or equal to zero for the first time, it needs to further verify that the absolute values ​​of R(n) for multiple consecutive sampling points (e.g., 10 points) following this point are all less than the rate of change stability judgment threshold. When both the zero-crossing condition and the condition of continuous subsidence below the threshold are met simultaneously, the current zero-crossing moment is determined as the dividing point between the rapid change phase and the equilibrium phase. The rapid change phase is defined as the time interval from the start of immersion to this dividing point.

[0074] The equilibrium stage begins at the aforementioned dividing point and continues until the end of the liquid level change sequence. In this stage, the absolute value of the liquid level change rate R(n) is generally less than the stability threshold, and its value alternates between positive and negative values, fluctuating around zero. This indicates that the net rate of change of liquid level has become very slow, and the system is in a state of dynamic adjustment around the final equilibrium position.

[0075] S4. Based on the recursive graph analysis method, determine whether the dynamic similarity between the driving control signal and the liquid level change sequence meets the requirements. The specific implementation is as follows:

[0076] First, phase space reconstruction is performed on both time series. Phase space reconstruction is achieved using the time delay method, which aims to expand the one-dimensional time series into a higher-dimensional space to reveal its underlying dynamic information. For the drive control signal sequence, a time delay parameter and an embedding dimension parameter need to be selected. The time delay parameter can be determined by calculating the number of delay steps corresponding to the decay of the autocorrelation function of the sequence to a specific proportion of its initial value. For example, the time delay can be calculated as the number of steps corresponding to the first decay of the autocorrelation function to 37% of its initial value. The embedding dimension parameter can be determined using the spurious nearest neighbor method. This method gradually increases the embedding dimension and calculates the proportion of spurious neighbors of adjacent points in the higher-dimensional space. When this proportion is lower than a preset tolerance, such as lower than 1%, the corresponding dimension can be selected as the embedding dimension. Applying the selected time delay and embedding dimension, each sampling point of the drive control signal sequence and its delayed points are combined into a multi-dimensional vector. This vector represents the phase space state point of the system at a certain moment. Connecting all state points in chronological order constitutes the phase space trajectory of the drive control signal. The liquid level change sequence was reconstructed independently using the same parameter selection principles and methods to obtain its corresponding phase space trajectory.

[0077] After obtaining the two phase space trajectories, their recursion matrices need to be calculated separately. The recursion matrix is ​​a two-dimensional square matrix, where the row and column indices correspond to the state point indices on the phase space trajectory. The core of calculating the recursion matrix is ​​setting a neighborhood radius parameter and determining the proximity relationship between any two state points. Setting the neighborhood radius is crucial for recursive analysis; its value aims to ensure that the proportion of recursive points (i.e., points with a value of one) in the final generated recursion matrix is ​​within a reasonable range. This range can be set to 1% to 10% to avoid the matrix becoming too sparse or too dense, thus losing its analytical significance. A specific setting method can be: first, calculate the average Euclidean distance between all pairwise state points in the corresponding phase space trajectory; then multiply this average by a coefficient, which can be chosen between 10% and 20%; through trial and error adjustments, finally select a value that ensures the proportion of recursive points falls within the aforementioned target range as the formal neighborhood radius. The specific process for calculating the recursion matrix of the phase space trajectory of the drive control signal is as follows: traverse each phase space state point in the trajectory and use it as a reference point; calculate the Euclidean distance between the reference point and all other state points in the phase space trajectory; compare each calculated Euclidean distance with the neighborhood radius set for the trajectory; if the Euclidean distance is less than or equal to the neighborhood radius, mark the intersection of the row corresponding to the reference point index and the column corresponding to the comparison point index in the recursion matrix with a value of one, indicating that recursion has occurred between these two state points in the phase space; if the Euclidean distance is greater than the neighborhood radius, mark the corresponding position with a value of zero, indicating that no recursion has occurred. After traversing all state points as reference points in this way, a complete drive control signal recursion matrix consisting of zero and one elements can be obtained. For the phase space trajectory of the liquid level change sequence, the same logic and steps are used to independently calculate and obtain its corresponding liquid level change recursion matrix, but its neighborhood radius parameter needs to be independently determined according to the distance distribution of its own trajectory.

[0078] After obtaining the two recursion matrices, recursive quantitative analysis needs to be performed on them separately to extract quantitative parameters characterizing the dynamic characteristics of their respective systems. Commonly used recursive quantitative analysis parameters include recursion rate, determinism, and stratification. Taking the recursion rate as an example, it is calculated as follows: the recursion rate equals the number of elements in the recursion matrix with a value of one divided by the total number of elements in the matrix. This parameter characterizes the overall probability that the system state returns to a neighboring region in phase space. When calculating the determinism parameter, all continuous line segments extending along the diagonal direction in the recursion matrix must first be identified. The determinism equals the total number of recursion points contained in these continuous line segments divided by the total number of recursion points in the matrix. This parameter characterizes the degree of determinism of the system's dynamic behavior. A set of identical parameters is calculated from both the drive control signal recursion matrix and the liquid level change recursion matrix, for example, by simultaneously calculating the recursion rate and determinism, thus forming the drive control signal parameter set and the liquid level change parameter set.

[0079] Finally, the similarity between the two parameter sets needs to be evaluated. A similarity metric is calculated between the two parameter sets; for example, Euclidean distance can be used. The similarity metric is calculated as follows: the similarity metric is equal to the sum of the squares of the differences between each pair of corresponding parameter values ​​in the two parameter sets, and then the square root of the sum. The smaller the similarity metric, the more similar the two sets of parameters are, meaning the dynamic characteristics of the two sequences are closer. The calculated similarity metric is compared with a preset dynamic similarity threshold. This threshold is set based on the level of similarity that two sequences should possess under numerous normal experimental conditions. Specifically, it can be obtained by conducting multiple standard immersion tests during the system debugging phase. For each test, the similarity metric between the drive control signal and the liquid level change sequence parameter set is calculated. The distribution of these metrics is then statistically analyzed, and the statistical average plus twice the standard deviation is taken as the dynamic similarity threshold. If the currently calculated similarity metric is less than or equal to the dynamic similarity threshold, the dynamic similarity is considered to meet the requirements; if it is greater than the threshold, the requirements are not met.

[0080] S5. If the requirements are met, extract the liquid level height data subset corresponding to the target stage used for stability determination in the dynamic stage, analyze the correlation pattern between the complexity and long-range correlation of the liquid level height data subset at multiple scales, and determine whether the liquid level is in a stable state that can be used for density calculation. The specific implementation is as follows:

[0081] First, from the dynamic stages identified in step S3, the equilibrium-reaching stage is selected as the target stage for stability determination. Specifically, based on the boundary point between the rapid change stage and the equilibrium-reaching stage determined in step S3, the portion of the liquid level change sequence from the boundary point to the end of the sequence is extracted. This subset of data is defined as the liquid level height data subset used for stability determination, representing the process of the liquid level change tending to equilibrium after the watermelon is submerged.

[0082] After obtaining the liquid level height data subset, multi-scale complexity and long-range correlation analysis needs to be performed on it. First, the multi-scale sample entropy of this data subset at different time scales is calculated to characterize its multi-scale complexity. The calculation of multi-scale sample entropy involves two steps: coarsening and sample entropy calculation. For a given time scale factor, for example, a scale factor ranging from 1 to 10, the original liquid level height data subset is coarsened. Specifically, the original data subset is divided into several non-overlapping segments, each with a length equal to the current scale factor. Then, the arithmetic mean of all data within each segment is calculated, and this mean is used to construct a new coarse-grained time series. Subsequently, the sample entropy of this coarse-grained series is calculated. The calculation of sample entropy requires setting two parameters: pattern dimension and tolerance threshold. The pattern dimension is typically set to 2, representing the length of the comparison vector. The tolerance threshold is used to determine whether two data patterns are similar; its specific value is set between 0.1 and 0.25 times the standard deviation of the current coarse-grained time series, for example, 0.2 times. This tolerance threshold is compared with the absolute value of the difference between corresponding elements of any two vectors calculated during the comparison process. If the absolute value of all corresponding differences is less than this tolerance threshold, the vector pair is considered similar. The specific calculation process of sample entropy is as follows: In the coarse-grained sequence, construct all vectors with a length equal to the pattern dimension; count the number of all vector pairs that satisfy the above similarity conditions; then increment the pattern dimension by 1 and repeat the above construction and statistical process; the sample entropy value is the negative natural logarithm of the ratio of the similar vector pair count result in the previous step to the count result in the next step. By repeating the above coarse-graining and sample entropy calculation process for each selected scale factor, a multi-scale sample entropy value sequence can be obtained, which reflects the change in the complexity of liquid level height fluctuations with the observation time scale.

[0083] Simultaneously, it is necessary to calculate the detrended volatility analysis scaling index for the same liquid level height data subset at the corresponding time scale to characterize its long-range correlation. The calculation steps for detrended volatility analysis are as follows: First, sum the liquid level height data subset cumulatively and subtract its arithmetic mean to generate a new profile sequence. Second, select a scale, which is the interval length, for example, starting with 10 data points, and not exceeding one-quarter of the total length of the data subset. Divide the profile sequence into multiple continuous, non-overlapping intervals of length equal to the current scale. Third, within each interval, fit a straight line using the least squares method as the local trend line for that interval. Fourth, within each interval, subtract the corresponding local trend line value from the original value of the profile sequence to obtain the detrended subsequence for that interval. Fifth, calculate the variance of the detrended subsequences in all intervals, then calculate the average of these variances, and finally take the square root of this average to obtain the volatility function value at the current scale. Sixth, change the scale value, for example, increasing it in powers of 2, and repeat steps two through five to obtain a set of volatility function values ​​at different scales. Step 7: In a log-log coordinate system, plot the data points with the scale value on the x-axis and the corresponding fluctuation function value on the y-axis. Use linear regression to fit a straight line to these points; the slope of the fitted line is the detrended fluctuation analysis scaling index for that subset of liquid level height data. For each coarse-grained scale used to calculate the multi-scale sample entropy, it needs to be consistent with the interval length concept and perform a complete detrended fluctuation analysis calculation independently, thus obtaining a detrended fluctuation analysis scaling index sequence corresponding to the scale of the multi-scale sample entropy sequence.

[0084] Next, we analyze the correlation pattern between the two sequences. The covariance between the multi-scale sample entropy sequence and the detrended volatility analysis scaling index sequence is calculated as a quantitative measure of the correlation pattern. The covariance is calculated as follows: First, the arithmetic mean of all values ​​in the multi-scale sample entropy sequence is calculated, and simultaneously, the arithmetic mean of all values ​​in the detrended volatility analysis scaling index sequence is calculated. Then, for each pair of corresponding values ​​in the sequences, the difference between each pair and the arithmetic mean of their respective sequences is calculated. Next, each pair of differences is multiplied. Finally, all products are summed and divided by the length of the sequence minus one. The resulting covariance value quantifies the degree of consistency in the changing trends of the two sequences. As the liquid level approaches physical equilibrium, complexity and long-range correlation should exhibit a specific cooperative changing pattern, and this covariance value is expected to fall within a specific numerical range.

[0085] Finally, the calculated correlation pattern, i.e., the covariance value, is compared with a preset stability criterion to determine whether the liquid level is in a stable state suitable for density calculation. This stability criterion is a preset covariance threshold range. This range is set based on the above-mentioned multi-scale analysis of a large amount of known liquid level height data in a stable equilibrium state, calculating the statistical distribution of its covariance values. Specifically, during the system calibration phase, multiple undisturbed, fully static standard immersion experiments are conducted. For each experiment, stable liquid level height data is collected, and the covariance value between the multi-scale sample entropy sequence and the detrended fluctuation analysis scaling index sequence is calculated independently for each data set. Then, the covariance values ​​obtained from all experiments are statistically analyzed, and the arithmetic mean and standard deviation of these covariance values ​​are calculated. The lower limit of the stability criterion can be set as the arithmetic mean minus twice the standard deviation, and the upper limit can be set as the arithmetic mean plus twice the standard deviation, thus forming a threshold range that includes both upper and lower limits. The comparison object of this covariance threshold range is the single covariance value calculated in the current experiment. The comparison method is as follows: determine whether the current covariance value is greater than or equal to the lower limit of the range, and simultaneously less than or equal to the upper limit of the range. If the covariance value calculated from the current experimental data falls within the preset covariance threshold range, the liquid surface is determined to be in a stable state that can be used for density calculation; if it falls outside the range, the liquid surface is determined to be in a stable state, and the data from this test cannot be used for subsequent density calculations.

[0086] S6. If the watermelon is in a stable state, determine the second water level based on the subset of water level data, and calculate the watermelon density by combining the first water level, the first weight, and the second weight. The specific implementation is as follows:

[0087] First, the second liquid level height required for calculation is determined. This step utilizes the valid liquid level height data subset extracted and determined in step S5. Statistical analysis is performed on all liquid level height values ​​contained in the liquid level height data subset, and their arithmetic mean is calculated. This arithmetic mean is then determined as the second liquid level height. The liquid level height data subset corresponds to the equilibrium stage identified in step S3. It has already passed the stability determination in step S5, indicating that the liquid level height fluctuates around a stable value within this time period. Therefore, using its arithmetic mean can effectively represent the final height after the watermelon is completely submerged and the liquid level reaches equilibrium. The method for calculating the arithmetic mean is: add the height values ​​of all sampling points in the liquid level height data subset, obtain the sum, and then divide by the total number of sampling points in the subset. Thus, the first liquid level height obtained in step S1 and the second liquid level height obtained at this point together form the basis for calculating the liquid level change.

[0088] After obtaining the second liquid level, the volume of water displaced when the watermelon is submerged needs to be calculated. This volume is obtained by multiplying the change in liquid level by the cross-sectional area of ​​the container. First, the change in liquid level is calculated. This change equals the second liquid level minus the first liquid level measured in step S1. The first liquid level is the stable liquid level when the supporting device is unloaded and submerged; the difference between this and the second liquid level when the watermelon is supported is the direct increase in liquid level caused by the water displaced by the watermelon. Next, the cross-sectional area of ​​the water-holding container needs to be obtained. For containers with a regular and uniform cross-sectional shape, the cross-sectional area is a fixed value. If the container is a regular cylinder, its cross-sectional area can be calculated by measuring or using the known inner diameter of the container. Specifically, using a length measuring tool, such as a digital vernier caliper, the inner diameter is measured in multiple directions near the container opening, and the arithmetic mean of the multiple measurements is taken as the inner diameter value for calculation. The expression for calculating the container's cross-sectional area is: cross-sectional area equals pi multiplied by the square of half the inner diameter. Pi can be taken as 3.1416. If the container's cross-sectional area has been precisely calibrated beforehand using other methods, the calibrated value can be used directly. Multiplying the calculated change in liquid level by the container's cross-sectional area will give the volume of water displaced by the watermelon. During the calculation, attention must be paid to the consistency and conversion of units. For example, the original unit for the change in liquid level is millimeters, which can be converted to centimeters; the unit for the cross-sectional area is square centimeters, and the unit for the displaced water volume is cubic centimeters, for subsequent density calculations.

[0089] Next, the weight of the watermelon itself is calculated. The weight of the watermelon is calculated using the two weight values ​​measured in step S1. In step S1, the first weight when the supporting device is unloaded, and the total weight of the supporting device and the watermelon together, i.e., the second weight, have been measured and stored. The net weight of the watermelon is equal to the second weight minus the first weight. This calculation is based on the linear characteristics of the weighing sensor; under the same measurement conditions, the difference between the two measurements is the accurate mass of the object placed on the supporting device, usually expressed in grams.

[0090] Finally, the density of the watermelon is calculated based on Archimedes' principle. When the watermelon is fully submerged and in a state of static equilibrium, its own weight is balanced by the buoyant force, therefore the mass of the watermelon is equal to the mass of the water it displaces. The mass of the displaced water is equal to the volume of the displaced water multiplied by the density of water. The density of water is a known physical constant. In practical applications, the corresponding density value can be adopted according to the measured water temperature. For example, under standard atmospheric pressure, the density of pure water at a temperature of 20 degrees Celsius is approximately 0.9982 grams per cubic centimeter. Based on this physical relationship, the calculation logic of the watermelon's density is expressed as: the density of the watermelon is equal to the mass of the watermelon divided by the volume of water displaced by the watermelon. Substituting the parameters obtained in the previous steps, the specific calculation is: MD = (L2 - L1) / ((H2 - H1) * S), where MD is the density of the watermelon, L2 is the second weight, L1 is the first weight, H2 is the second liquid level height, H1 is the first liquid level height, and S is the bottom area of ​​the water tank. To ensure the correct physical meaning and consistent units of the calculation results, it is recommended to convert all parameters to a system using grams as the unit of mass and centimeters as the unit of length for calculation. During the calculation, the density of water is used as a constant and is actually canceled out in the formula derivation. Therefore, the final calculation relies only on directly measured weight differences, liquid level differences, and geometric dimensions, without needing to explicitly use the density value of water. This eliminates errors that may be introduced by changes in water temperature.

[0091] Example 2: Figure 2 A schematic diagram of the watermelon maturity non-destructive testing system based on the water immersion density method of the present invention is given. The watermelon maturity non-destructive testing system based on the water immersion density method includes:

[0092] The data acquisition module is used to acquire the first liquid level height and first weight of the carrier when it is unloaded and submerged in water, and the second weight of the carrier after it is loaded with watermelon.

[0093] The sequence formation module is used to control the carrier device to immerse the watermelon in water, synchronously record the drive control signal and collect the liquid level height to form a sequence of liquid level height changes;

[0094] The phase identification module is used to identify the dynamic phases in the watermelon immersion process based on the sequence of liquid level changes;

[0095] The requirement is to have a judgment module that, based on a recursive graph analysis method, determines whether the dynamic similarity between the driving control signal and the sequence of liquid level changes meets the requirements.

[0096] The state judgment module is used to extract the liquid level height data subset corresponding to the target stage for stability judgment in the dynamic stage if the requirements are met, analyze the correlation pattern between the complexity and long-range correlation of the liquid level height data subset at multiple scales, and determine whether the liquid level is in a stable state that can be used for density calculation.

[0097] The density calculation module is used to determine the second liquid level height based on a subset of liquid level height data if the liquid level is in a stable state, and to calculate the watermelon density by combining the first liquid level height, the first weight, and the second weight.

[0098] The calculations involved in the embodiments are all dimensionless numerical calculations, and the preset parameters and thresholds in the calculations are set by those skilled in the art according to the actual situation.

[0099] It should be noted that this invention can be deployed on the device itself to realize embedded applications, or it can run on a PC or other terminal with a user interface, thereby meeting various hardware environments and usage requirements.

[0100] The above embodiments can be implemented, in whole or in part, by software, hardware, firmware, or any other combination thereof. When implemented using software, the above embodiments can be implemented, in whole or in part, as a computer program product. A computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, all or part of the processes or functions according to the embodiments of this application are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. Computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wireless or wired transmission; wired transmission methods include optical fiber, twisted pair, coaxial cable, etc.; wireless transmission includes infrared, microwave, etc. Computer-readable storage media can be any available medium that a computer can access or a data storage device such as a server or data center that contains one or more sets of available media. Available media can be magnetic media (e.g., floppy disks, hard disks, magnetic tapes), optical media (e.g., DVDs), or semiconductor media. Semiconductor media can be solid-state drives.

[0101] Those skilled in the art will understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and modules described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0102] In the several embodiments provided in this application, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of modules is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple modules or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between apparatuses or modules may be electrical, mechanical, or other forms.

[0103] The modules described as separate components may or may not be physically separate. The components shown as modules may or may not be physical modules; they may be located in one place or distributed across multiple network modules. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs.

[0104] In addition, the functional modules in the various embodiments of this application can be integrated into one processing module, or each module can exist physically separately, or two or more modules can be integrated into one module.

[0105] If a function is implemented as a software module and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0106] The above are merely specific embodiments of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

[0107] In conclusion, the above are merely preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A non-destructive method for detecting watermelon maturity based on the water immersion density method, characterized in that, include: S1. Obtain the first liquid level and first weight of the bearing device when it is unloaded and submerged in water, and the second weight of the bearing device after it is loaded with watermelon; S2. Control the carrier device to immerse the watermelon in water, synchronously record the drive control signal and collect the liquid level height to form a sequence of liquid level height changes; S3. Based on the sequence of changes in liquid level, identify the dynamic stages in the process of immersing watermelons; S4. Based on the recursive graph analysis method, determine whether the dynamic similarity between the driving control signal and the liquid level change sequence meets the requirements, including: Phase space reconstruction was performed on the driving control signal and the liquid level change sequence to obtain the phase space trajectories corresponding to the two time series respectively. Each recursion matrix is ​​calculated based on the phase space trajectory. The elements in the recursion matrix indicate whether the phase space state point at the corresponding time is within the set neighborhood radius. Recursive quantitative analysis is performed on the two recursive matrices to extract parameters characterizing the dynamic properties of the system. Calculate the similarity metric between two parameter sets, compare the similarity metric with a preset threshold, and determine whether the dynamic similarity meets the requirements based on the comparison result; S5. If the requirements are met, extract the liquid level height data subset corresponding to the target stage used for stability determination in the dynamic stage, analyze the correlation pattern between the complexity and long-range correlation of the liquid level height data subset at multiple scales, and determine whether the liquid level is in a stable state that can be used for density calculation. S6. If the watermelon is in a stable state, determine the second water level based on the subset of water level data, and calculate the watermelon density by combining the first water level, the first weight, and the second weight.

2. The non-destructive testing method for watermelon maturity based on the water immersion density method according to claim 1, characterized in that, S1 includes: When the load-bearing device is in an unloaded state, the load-bearing device is controlled to be completely submerged in water. The first liquid level height is measured by a liquid level sensor, and the first weight of the unloaded load-bearing device is measured by a weighing sensor. Next, the watermelon was loaded onto the support device, and the combined weight of the support device and the watermelon was measured using the same weighing sensor.

3. The non-destructive testing method for watermelon maturity based on the immersion density method according to claim 1, characterized in that, S2 include: The control device carries the watermelon into the water to perform the immersion operation. During this process, the drive control signal that drives the movement of the device is recorded synchronously and continuously, and the liquid level height is collected at a fixed sampling frequency by the liquid level sensor. From the moment the watermelon begins to touch the water surface, data is continuously recorded and collected until the immersion process is completed and the predetermined collection time is reached. The collected water level data, arranged in chronological order, forms a sequence of water level changes.

4. The non-destructive testing method for watermelon maturity based on the immersion density method according to claim 1, characterized in that, S3 include: Differential calculations are performed on the sequence of liquid level changes to obtain the corresponding sequence of liquid level change rates; Based on the numerical distribution and zero-crossing characteristics of the liquid level height change rate sequence, the watermelon immersion process is divided into a dynamic stage that includes at least a rapid change stage and a tendency to equilibrium stage.

5. The non-destructive testing method for watermelon maturity based on the immersion density method according to claim 1, characterized in that, The calculation of the recursion matrix based on the phase space trajectory includes: for each phase space state point in the phase space trajectory, calculating the Euclidean distance between the corresponding phase space state point and all other phase space state points in the phase space trajectory, and comparing the calculated Euclidean distance with a pre-set neighborhood radius; if a certain Euclidean distance is less than or equal to the corresponding neighborhood radius, the corresponding position in the recursion matrix is ​​marked as the first preset value, indicating that recursion has occurred between the two corresponding phase space state points; if it is greater than the corresponding neighborhood radius, it is marked as the second preset value, indicating that no recursion has occurred; this process is repeated for all phase space state points to form a complete recursion matrix.

6. The non-destructive testing method for watermelon maturity based on the immersion density method according to claim 1, characterized in that, S5 include: From the identified dynamic stages, the equilibrium stage is selected as the target stage for stability determination, and the corresponding liquid level height data subset is extracted for the target stage. The multi-scale sample entropy of liquid level height data subsets at different time scales is calculated to characterize the complexity at multiple scales. At the same time, the detrended fluctuation analysis scaling index of the corresponding liquid level height data subset at the corresponding scale is calculated to characterize the long-range correlation. The covariance between the multi-scale sample entropy sequence and the detrended fluctuation analysis scaling exponential sequence is analyzed as a correlation pattern. The correlation pattern is compared with a preset stability criterion to determine whether the liquid surface is in a stable state that can be used for density calculation.

7. The non-destructive testing method for watermelon maturity based on the water immersion density method according to claim 6, characterized in that, The calculation of the detrended volatility analysis scaling index at the corresponding scale for a subset of liquid level height data to characterize long-range correlation includes: cumulatively summing and removing the mean from the subset of liquid level height data to generate a new profile sequence; dividing the profile sequence into multiple non-overlapping intervals of equal length, fitting the local trend within each interval using the least squares method to obtain the local trend, and subtracting the local trend from the profile sequence to obtain the detrended subsequence; calculating the average variance of the detrended subsequences within all intervals, and then taking the square root to obtain the volatility function value at the corresponding scale; changing the interval length to correspond to different time scales to obtain volatility function value sequences at different scales; and linearly fitting the volatility function value sequence to the corresponding scale in a log-log coordinate system, with the slope of the fitted line being the detrended volatility analysis scaling index.

8. The non-destructive testing method for watermelon maturity based on the immersion density method according to claim 1, characterized in that, S6 include: Statistical analysis was performed on a subset of liquid level data, and the average liquid level was calculated as the second liquid level. The change in water level caused by the immersion of the watermelon is calculated based on the difference between the first and second water level heights, and the volume of water displaced by the watermelon is calculated based on the known cross-sectional area of ​​the container. Calculate the weight of the watermelon based on the first and second weights; Based on Archimedes' principle, the density of a watermelon can be calculated using its weight and the volume of water it displaces.

9. A non-destructive testing system for watermelon maturity based on the water immersion density method, used to implement the non-destructive testing method for watermelon maturity based on the water immersion density method as described in any one of claims 1-8, characterized in that, include: The data acquisition module is used to acquire the first liquid level height and first weight of the carrier when it is unloaded and submerged in water, and the second weight of the carrier after it is loaded with watermelon. The sequence formation module is used to control the carrier device to immerse the watermelon in water, synchronously record the drive control signal and collect the liquid level height to form a sequence of liquid level height changes; The phase identification module is used to identify the dynamic phases in the watermelon immersion process based on the sequence of liquid level changes; The requirement is to have a judgment module that, based on a recursive graph analysis method, determines whether the dynamic similarity between the driving control signal and the sequence of liquid level changes meets the requirements. The state judgment module is used to extract the liquid level height data subset corresponding to the target stage for stability judgment in the dynamic stage if the requirements are met, analyze the correlation pattern between the complexity and long-range correlation of the liquid level height data subset at multiple scales, and determine whether the liquid level is in a stable state that can be used for density calculation. The density calculation module is used to determine the second liquid level height based on a subset of liquid level height data if the liquid level is in a stable state, and to calculate the watermelon density by combining the first liquid level height, the first weight, and the second weight.

Images