A method for improving the content of bird's nest peptides

Abstract

The present application relates to the technical field of bird's nest peptide preparation, and discloses a method for assisting in improving the peptide content of bird's nest peptide. The method comprises the following steps: obtaining a multi-parameter original data set collected by a monitoring device within a preset historical period, the multi-parameter original data set comprising historical temperature, humidity and time data sets; performing overall stability calculation on the data set to derive temperature, humidity and time stability values; taking the stability values as guide parameters to perform intensive scanning in a predefined bird's nest processing parameter space to locate an optimal processing period; performing peptide content optimization processing on the bird's nest raw material according to the period to generate a serialized processing data set, calling a peptide content determination unit for evaluation, and outputting a bird's nest peptide content improvement evaluation result. The method determines processing conditions based on historical multi-parameters and stability analysis, forms a systematic scheme in combination with serialized evaluation, adapts to different raw material characteristics, and improves the stability and reliability of processing effect.

Description

Technical Field

[0001] This invention relates to the field of bird's nest peptide preparation technology, specifically a method to help increase the peptide content of bird's nest. Background Technology

[0002] Bird's nest peptides, small-molecule bioactive peptides extracted from bird's nest, are increasingly widely used in health products and functional foods due to their rich amino acid composition and good bioactivity. The industry's demand for high-content bird's nest peptide products is also constantly rising. However, current techniques for increasing peptide content in the preparation of bird's nest peptides still have many limitations, making it difficult to achieve stable and efficient peptide content optimization.

[0003] In existing technologies, most methods for increasing the peptide content of bird's nest rely on empirical parameter settings, lacking a systematic consideration of the synergistic effects of multiple factors. For example, some solutions adjust processing conditions solely based on real-time monitoring of a single temperature parameter, ignoring the temperature variation patterns of bird's nest raw materials over a predetermined historical period of long-term storage and pretreatment. If the historical temperature data of a batch of bird's nest raw materials fluctuates frequently, determining processing parameters solely based on achieving the real-time temperature standard can easily lead to uneven peptide bond breakage during subsequent processing, thus affecting the final peptide content. Simultaneously, existing technologies do not pay sufficient attention to humidity parameters, mostly only controlling humidity within a broad acceptable range without analyzing the correlation between historical humidity data and temperature and time parameters. Bird's nest raw materials have strong hygroscopic properties; unstable humidity can lead to clumping or excessive drying, damaging the internal structure of the raw materials and hindering the effective dissolution of peptides.

[0004] In determining the processing cycle, existing technologies mostly rely on limited comparative experiments or subjective experience of operators, lacking scientific guidance and comprehensive parameter coverage. Some schemes only select a few fixed processing cycle durations for experiments, comparing peptide content under different cycles to determine the "optimal cycle." However, this approach covers a limited combination of parameters and cannot adapt to bird's nest raw materials with different historical parameter characteristics. When changing batches of raw materials, the original "optimal cycle" may no longer be applicable, requiring numerous new experiments, which not only increases time and cost but also makes it difficult to guarantee the consistency of processing results.

[0005] In the peptide content assessment stage, existing technologies generally only perform a single measurement on the final processed product, failing to dynamically track data during the processing. This assessment method cannot promptly detect abnormal changes in peptide content at different stages of processing. For example, if peptide content growth has stagnated in the middle of processing, continuing processing according to the original cycle will result in a waste of raw materials and energy. Furthermore, the lack of process data accumulation cannot provide effective references for optimizing subsequent processing parameters, leading to slow technological iteration and upgrades, and making it difficult to meet the industry's demand for large-scale production of high-content bird's nest peptide products. Summary of the Invention

[0006] The purpose of this invention is to provide a method to help increase the peptide content of bird's nest, so as to solve the problems mentioned in the background art.

[0007] To achieve the above objectives, the present invention provides a method for assisting in increasing the peptide content of bird's nest, the method comprising:

[0008] The raw data set of bird's nest materials is collected by monitoring equipment within a preset historical period. The raw data set of the multi-parameter data includes historical temperature dataset, historical humidity dataset and historical time dataset.

[0009] The overall stability of the multi-parameter raw data set is calculated, and the values ​​of temperature stability, humidity stability and time stability are derived.

[0010] Using the temperature stability value, humidity stability value, and time stability value as guiding parameters, a dense scanning operation is performed within the predefined bird's nest processing parameter space to locate the optimal processing cycle.

[0011] Based on the optimal processing cycle, the bird's nest raw materials are subjected to peptide content optimization treatment, a serialized processing data set is generated, and the peptide content determination unit is called to evaluate the serialized processing data set, and the evaluation result of the improvement of bird's nest peptide content is output.

[0012] Preferably, acquiring the multi-parameter raw data set includes: deploying distributed sensing devices to continuously capture temperature change signals, humidity change signals, and time elapsed signals in the bird's nest processing environment, and archiving them into the historical temperature dataset, historical humidity dataset, and historical time dataset according to a preset sampling frequency.

[0013] Preferably, the overall stability calculation of the multi-parameter raw data set includes: calculating the average value and variation within the moving window for the historical temperature dataset, historical humidity dataset, and historical time dataset respectively, and merging the average value and variation to generate the temperature stability value, humidity stability value, and time stability value.

[0014] Preferably, calculating the average and variation within the moving window includes: setting a sliding time window of fixed length, and calculating the average temperature and temperature variation of the historical temperature dataset, the average humidity and humidity variation of the historical humidity dataset, and the average time and time variation of the historical time dataset window by window.

[0015] Preferably, performing intensive scanning operations within a predefined bird's nest processing parameter space includes: establishing a three-dimensional parameter space with temperature stability values ​​as the horizontal axis, humidity stability values ​​as the vertical axis, and time stability values ​​as the vertical axis; and loading multiple sample stability values ​​and corresponding sample processing cycles as basic data points in the three-dimensional parameter space.

[0016] Preferably, determining the optimal processing cycle includes: using the current temperature stability value, humidity stability value, and time stability value as reference points, delineating a spherical search region in the three-dimensional parameter space, and calculating the distribution density of basic data points within the spherical search region.

[0017] Preferably, calculating the distribution density includes: constructing a spherical subspace with the reference point as the center and a preset search radius as the scale; counting the number of basic data points contained in the spherical subspace, and dividing the number of basic data points by the volume of the spherical subspace to obtain the local density value.

[0018] Preferably, the optimal processing cycle for locating the target point further includes: randomly selecting a basic data point from the boundary of the spherical search region as a candidate point, calculating the neighborhood density of the candidate point; if the neighborhood density is greater than or equal to the current local density value, updating the candidate point as a new reference point, repeating the density comparison process until the maximum number of iterations is reached, and taking the sample processing cycle corresponding to the final reference point as the optimal processing cycle.

[0019] Preferably, the process of optimizing the peptide content of bird's nest raw materials according to the optimal processing cycle includes: adjusting the heating temperature, moisturizing level and processing time of bird's nest according to the optimal processing cycle, performing multiple rounds of peptide extraction operations, and recording the peptide concentration data after each round of operations to form the serialized processing data set.

[0020] Preferably, the evaluation by calling the peptide content determination unit includes: using spectral analysis technology to verify the peptide concentration data in the serialized data set point by point, calculating the average peptide content and deviation value, and generating the evaluation result of the improvement of bird's nest peptide content based on the average peptide content and deviation value.

[0021] Compared with the prior art, the beneficial effects of the present invention are:

[0022] This design acquires a set of raw data on multiple parameters of bird's nest raw materials within a preset historical period, including historical temperature, humidity, and time datasets. This approach breaks away from the limitations of existing technologies that rely on single, real-time data. It fully utilizes the historical parameter characteristics of the raw materials. Different batches and sources of bird's nest raw materials exhibit differences in environmental parameters during storage and pretreatment. This historical data reflects the raw materials' adaptability to different parameters, providing fundamental information that aligns with the actual conditions of the raw materials for subsequent processing parameter optimization. This avoids processing deviations caused by mismatches between parameter settings and the historical adaptability characteristics of the raw materials, making parameter settings more targeted.

[0023] After acquiring raw data for multiple parameters, an overall stability calculation is performed, and the stability values ​​for temperature, humidity, and time are derived, transforming scattered parameter information into quantifiable objective indicators. Traditional methods for judging parameter stability often remain at a subjective level or rely on simple range determinations, failing to accurately reflect the details of parameter fluctuations. This method, however, can demonstrate the fluctuation amplitude and duration of sustained stability of parameters within a cycle, while simultaneously considering the stability of all three parameters, rather than a single parameter. For example, even if both temperature and humidity are within acceptable ranges, significant differences in their stability can still affect the raw material processing effect. Overall stability calculation avoids the limitations of single-parameter assessments, providing scientific guidance for determining the optimal processing cycle.

[0024] Guided by stability values, this method performs intensive scanning within a predefined processing parameter space to pinpoint the optimal processing cycle, further improving the accuracy and efficiency of cycle determination. Existing technologies suffer from narrow cycle screening coverage and weak targeting, while this method, through a predefined parameter space, ensures that scanning always revolves around a reasonable range for bird's nest processing, avoiding blind scanning. Simultaneously, guided by stability values, intensive scanning focuses on parameter combinations that match the historical stability characteristics of the raw materials, significantly improving scanning effectiveness. It can accurately identify cycles conducive to peptide content enhancement within a wide range of parameter combinations, avoiding excessively long cycles leading to over-processing of the raw materials or excessively short cycles resulting in insufficient peptide dissolution, thus balancing effectiveness and resource conservation.

[0025] By processing raw materials according to the optimal cycle and generating a serialized processing dataset, and then evaluating this dataset using a measurement unit, dynamic control of the processing process is achieved. Existing technologies only measure the final product and cannot detect process anomalies in a timely manner. However, the serialized dataset records peptide content changes at each stage of processing. Evaluation allows for real-time monitoring of processing progress, timely identification of anomalies and adjustments, and reduction of ineffective processing. Simultaneously, the serialized data enriches the multi-parameter raw dataset, providing a reference for parameter optimization of subsequent batches of raw materials, forming a data-driven iterative mechanism that continuously improves the method's adaptability and processing effectiveness with data accumulation. Attached Figure Description

[0026] Figure 1 This is a schematic diagram illustrating the working principle of the method for assisting in increasing the peptide content of bird's nest according to the present invention.

[0027] Figure 2 A flowchart for obtaining the original data set with multiple parameters;

[0028] Figure 3 A flowchart for calculating the mean and variation within a moving window. Detailed Implementation

[0029] 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.

[0030] Please see Figure 1 This invention provides a method to assist in improving the peptide content of bird's nest. The method includes: acquiring a set of multi-parameter raw data collected by a monitoring device from bird's nest raw materials within a preset historical period, the set of multi-parameter raw data including historical temperature dataset, historical humidity dataset, and historical time dataset; performing an overall stability calculation on the set of multi-parameter raw data to derive temperature stability values, humidity stability values, and time stability values; using the temperature stability values, humidity stability values, and time stability values ​​as guiding parameters, performing a dense scanning operation within a predefined bird's nest processing parameter space to locate the optimal processing period; applying peptide content optimization processing to the bird's nest raw materials according to the optimal processing period, generating a serialized processed data set, and calling a peptide content determination unit to evaluate the serialized processed data set, outputting an evaluation result of the improvement in bird's nest peptide content.

[0031] Example 1: See Figure 2The collection of multi-parameter raw data relies on the deployment of a distributed sensing device. This device consists of a series of high-precision environmental sensor nodes strategically installed in various physical locations within the bird's nest processing environment, including corners of the raw material storage area, the space above the washing tank, near the inlet and outlet of the cooking equipment, and within the airflow channels of the drying chamber. Each node of the distributed sensing device integrates a temperature sensing module, a humidity sensing module, and a precision timing module. The temperature sensing module uses a platinum resistance temperature detector, with a measurement range covering -10°C to 100°C. The humidity sensing module uses a polymer thin-film capacitive humidity sensor, capable of detecting relative humidity from 5% to 99%. The precision timing module is driven by a crystal oscillator, providing millisecond-level timestamp recording capability. The distributed sensing device immediately enters continuous monitoring mode upon power-up. Temperature change signals are output as voltage from the temperature sensing module, humidity change signals are represented as changes in capacitance, and time elapsed signals are generated by pulse counting from the precision timing module. These raw signals are amplified and filtered by the signal conditioning circuit inside the node to eliminate power frequency interference and random noise. Then, they are converted into digital quantities by an analog-to-digital converter. The digital quantities are transmitted to the central data logger through the wireless communication unit of the distributed sensing device. The central data logger archives the received data according to a preset sampling frequency, which is set to collect data points once per second. The archiving operation arranges the temperature digital quantities in chronological order to form a historical temperature dataset, serializes the humidity digital quantities into a historical humidity dataset, and stores the timestamp information as a historical time dataset. The historical temperature dataset adopts a two-dimensional array structure, with the first column being the Unix timestamp and the second column being the corresponding Celsius temperature value. The historical humidity dataset has the same structure but stores the percentage humidity value. The historical time dataset records the start time and duration of each sampling interval.

[0032] Overall stability testing of multi-parameter raw datasets is a computationally intensive task. The overall stability test is performed separately for historical temperature, humidity, and time datasets. The calculation process is based on moving window statistical theory, with a fixed window length of one hundred consecutive sampling points, corresponding to a real-world time span of one hundred seconds. The moving window starts covering the index zero position of each dataset. The average temperature value within the window is calculated as the arithmetic mean of the one hundred temperature values. Simultaneously, temperature variability is calculated, quantified using the sample standard deviation formula, which is expressed as the square root of the sum of the squares of the differences between each temperature value and the average temperature value, divided by ninety-nine. The average humidity value and humidity variability for the historical humidity dataset are calculated using the same method. The average humidity value is the mean of humidity readings within the window, and the humidity variability reflects the degree of fluctuation of humidity values ​​around the average humidity value. The calculation of the time average value for the historical time dataset is slightly different. The time average value is the average duration of one hundred time intervals within the window, and the time variability is characterized by the standard deviation of these durations. The moving window slides forward one sampling point after completing the calculation at one location. The window slides using a first-in, first-out (FIFO) queue mechanism; new data points are added from the tail of the queue, and the oldest data points are removed from the head, ensuring the window always contains one hundred of the most recent data points. The fusion of the mean and variability is achieved through a linear weighted combination. The temperature stability value is obtained by multiplying the temperature mean by a weighting factor of 0.6 and adding the temperature variability by a weighting factor of 0.4. The humidity stability value uses the same weighting to fuse the humidity mean and humidity variability. The time stability value is generated by using a weight of 0.7 for the time mean and 0.3 for the time variability. The weighted temperature stability, humidity stability, and time stability values ​​are written to a stability report file in real time. The report file uses a comma-separated value format, with each line containing the window start timestamp and the corresponding three stability values. These values ​​serve as guiding parameters for subsequent steps.

[0033] There is a close coupling between the data flow of the distributed sensing device and the overall stability measurement process. The continuity of the data flow directly affects the accuracy of the statistics within the moving window. The nodes of the distributed sensing device use a communication protocol with forward error correction when transmitting data, effectively avoiding incomplete window data caused by packet loss. The central data logger performs integrity checks when receiving data; any data frame that fails the check will request retransmission from the sending node, ensuring that the historical temperature, humidity, and time datasets are uninterrupted or do not jump. The moving window calculation module resides in a high-performance embedded processor equipped with a floating-point unit to accelerate the calculation of the mean and standard deviation. The sliding mechanism of the moving window is implemented through a circular buffer; the buffer pointer automatically wraps back to the starting address when it reaches the end of the array. This design reduces memory copy operations and improves computational efficiency. The generation of the stability report file is asynchronous. A separate log thread is responsible for writing the calculated temperature stability, humidity stability, and time stability values ​​into non-volatile memory. The memory sector management uses a wear-leveling algorithm to extend the lifespan of the storage device.

[0034] Example 2: See Figure 3 The calculation process is based on a fixed-length sliding time window, with the window length set to one hundred consecutive sampling points. The sliding time window is initialized starting with the initial indices of the historical temperature dataset, historical humidity dataset, and historical time dataset. The historical temperature dataset contains temperature samples arranged in chronological order, the historical humidity dataset stores the corresponding humidity measurement sequences, and the historical time dataset records the timestamp information for each sampling point. The sliding time window is implemented in memory as a circular buffer. The pointer of the circular buffer points to the current starting position of the window, and the size of the buffer is fixed at one hundred data points, precisely accommodating a complete instance of the sliding time window. When a new data point arrives from the distributed sensing device, the new data point is written to the current position of the end pointer in the circular buffer, while the oldest data point is overwritten from the head of the buffer. The sliding time window moves forward through this mechanism, with the step size strictly maintained at one sampling point interval. The circular buffer design avoids large-scale data copy operations, improving computational efficiency, and the pointer automatically wraps back to the starting address when it reaches the end of the buffer.

[0035] For the historical temperature dataset, the average temperature within each sliding time window is calculated by summing one hundred temperature samples within the window and dividing by one hundred. The summation operation is implemented using an accumulator register, which employs an incremental update strategy as the window slides. This incremental update strategy subtracts the oldest covered temperature value from the accumulator and adds the newly arrived temperature value as the window moves, avoiding recalculating the sum for the entire window each time. Temperature variability is measured using standard deviation. The standard deviation is calculated by first determining the average temperature, then calculating the sum of squared deviations of each temperature value from the average. Dividing the sum of squared deviations by ninety-nine yields the variance, and the positive square root of the variance is the temperature variability. The calculation of the sum of squared deviations also employs an optimization strategy, maintaining a sum of squared deviations accumulator, which is dynamically updated based on old and new values ​​as the window slides. The calculation of the average humidity and humidity variability in the historical humidity dataset follows the same mathematical principles and calculation strategies as the historical temperature dataset. The average humidity is the arithmetic mean of the humidity values ​​within the window, and the humidity variability is the standard deviation of these humidity values. The time average of historical time datasets is calculated based on the time interval between consecutive sampling points within a window. The time average is the mean of these time intervals, while the time variability is the standard deviation of the time interval sequence, reflecting the jitter of the time signal.

[0036] The sliding time window movement and calculation process are synchronized. Each sampling period triggers one window sliding and one complete calculation. The sampling period is determined by the preset sampling frequency of the distributed sensing device, which is one hertz. Therefore, the sliding time window moves forward one position every second. Immediately after the movement, a new calculation period is triggered. This period includes reading all data points of the current window from the circular buffer and performing numerical calculations on the average temperature, temperature variability, average humidity, humidity variability, average time, and time variability. These numerical calculations are performed by the floating-point unit of the embedded system, which provides hardware-accelerated addition, multiplication, and square root operations to ensure that the calculations are completed within the sampling interval. The calculation results are written to a temporary structure containing the window's start timestamp and the six calculated statistics. This temporary structure is then passed to the data fusion module for fusion of stability values. The entire calculation process is designed as a non-blocking real-time task, with the calculation task having a higher interrupt priority than the data acquisition task to prevent data loss or calculation delays.

[0037] The circular buffer's memory management employs Direct Memory Access (DMI) technology. The DMI controller automatically transfers data from the input / output ports to the designated location in the circular buffer when sensor data arrives. DMI transfers do not require CPU intervention, freeing up CPU computing resources for statistical operations. The circular buffer's index management is implemented using two pointers: a write pointer indicates the location of the next data write, and a read pointer indicates the starting position of the current sliding time window. The write pointer always leads the read pointer, with the difference between them maintained at the length of the sliding time window (one hundred). When the write pointer reaches the end of the buffer, it automatically wraps back to the starting address; the read pointer's movement also follows this wraparound rule. Pointer comparison and update operations are performed in interrupt service routines, which are triggered each time sensor data arrives, ensuring atomic modification of the pointer state. This design guarantees the correct sliding of the sliding time window over the continuous data stream, avoiding misalignment or omission of window data. The average temperature, temperature variability, average humidity, humidity variability, time average, and time variability involved in the calculation are all intermediate results. These intermediate results are used to generate temperature stability, humidity stability, and time stability values. The temperature stability value is obtained by multiplying the average temperature by a weighting coefficient of 0.6 and the temperature variation by a weighting coefficient of 0.4, and then adding the results. Similarly, the humidity stability value is obtained by multiplying the average humidity by 0.6 and the humidity variation by 0.4, and the time stability value is obtained by multiplying the average time by 0.7 and the time variation by 0.3. The weighting coefficients are stored in non-volatile memory and can be adjusted according to different bird's nest processing requirements. The weighted calculation is performed immediately after the statistical calculation is completed. The result of the weighted calculation, as the final output of the stability measurement, is written to the system log and sent to the parameter space scanning module.

[0038] Example 3: The predefined bird's nest processing parameter space is a three-dimensional mathematical structure. The horizontal axis of the three-dimensional parameter space is defined as the temperature stability value, the vertical axis as the humidity stability value, and the time stability value. The temperature stability value, humidity stability value, and time stability value are all derived from the output results of the previous stability calculation steps. The coordinate range of the three-dimensional parameter space is dynamically determined by the minimum and maximum stability values ​​in the historical dataset. For example, the temperature stability value ranges from 0.1 to 0.9, the humidity stability value ranges from 0.2 to 0.8, and the time stability value ranges from 0.3 to 0.7. The coordinate axes use a linear scale, and each unit interval on each axis represents a stability change of 0.1. The three-dimensional parameter space is implemented in computer memory as a virtual coordinate system. The coordinate system uses floating-point numbers to represent point positions, with precision retained to four decimal places to ensure the accuracy of numerical calculations. Multiple sample stability values ​​and corresponding sample processing cycles are loaded into the three-dimensional parameter space as basic data points. The sample stability values ​​are derived from archived bird's nest processing experimental records. Each sample stability value includes a temperature stability value, a humidity stability value, and a time stability value. The sample processing cycle records the processing time parameters under the corresponding experimental conditions, such as heating duration, moisturizing cycle, and total processing time. The number of basic data points is typically set between several hundred and several thousand to cover most areas of the parameter space. The basic data points are stored in array form, with each point containing three-dimensional coordinates and additional processing cycle information. The spacing between data points is optimized using a spatial index structure such as an octree to accelerate subsequent search operations.

[0039] The dense scanning operation starts with the current temperature stability, humidity stability, and time stability values, which together form a reference point. This reference point serves as the initial search position input into the three-dimensional parameter space. The core of the dense scanning operation is to define a spherical search region within the three-dimensional parameter space. This region is centered on the reference point, and its radius is controlled by a preset search radius parameter. This preset radius is adaptively adjusted based on the overall size of the parameter space, for example, set to a stability unit of 0.15. The spherical search region is defined by calculating the Euclidean distance between all points and the reference point. The Euclidean distance formula, used to determine if a point is within the sphere, is expressed as the square root of the sum of the squared differences in the x-coordinate, y-coordinate, and vertical coordinates. The basic data points within the spherical search region are extracted for distribution density calculation. This calculation is a crucial step in determining the optimal processing cycle, reflecting the concentration of data points near the reference point.

[0040] The calculation of distribution density introduces a mathematical formula to quantify the point distribution within a spherical search region. The formula is defined as: distribution density equals the number of basic data points contained within the spherical search region divided by the volume of the spherical search region. The meanings of all characters in the formula are as follows: distribution density represents the number of basic data points per unit volume, which is a scalar value; the number of basic data points is the total number of data points within the spherical search region that satisfy the distance condition, which is an integer; the volume of the spherical search region is the volume of a sphere, calculated based on the sphere's radius. The mathematical expression of the formula is:

[0041]

[0042] in: Represents the distribution density. Represents the number of basic data points. The volume represents the spherical search region, and the formula for calculating the volume is: ,in It is the radius of the spherical search region.

[0043] The optimal processing cycle is determined based on the comparison and iterative optimization of distribution density. The process begins at a reference point, calculating the distribution density of the current spherical search area, denoted as the current local density value. Then, a baseline data point is randomly selected from the boundary of the spherical search area as a candidate point. The selection of candidate points uses a uniform random sampling algorithm to ensure that all boundary points have an equal probability of being selected. The neighborhood density of the candidate point is calculated in the same way as the distribution density, but a new spherical subspace is constructed with the candidate point as the new center. The radius of the new spherical subspace is the same as the radius of the original spherical search area. Neighborhood density represents the concentration of points near the candidate point. If the neighborhood density of a candidate point is greater than or equal to the current local density value, the candidate point is updated to the new reference point, and the density comparison process is repeated. This process continues until the maximum number of iterations is reached. The maximum number of iterations is set to a fixed value, such as one hundred, to prevent infinite loops. The final processing cycle corresponding to the reference point is output as the optimal processing cycle. The optimal processing cycle includes specific temperature, humidity, and time parameter settings to guide subsequent bird's nest processing operations.

[0044] The establishment of the 3D parameter space and the implementation of dense scanning operations rely on high-performance computing hardware. The data structure of the 3D parameter space is stored using a 3D array or a point cloud database. The point cloud database supports fast range queries and neighborhood searches. The calculation of the spherical search region is accelerated through spatial partitioning algorithms, such as using a grid index to divide the space into small cubic cells, checking only the points within the cells intersecting the sphere, reducing unnecessary distance calculations. The volume calculation in the distribution density formula pre-caches volume values ​​under common radii to improve computational efficiency, and dynamically updates the cache when the radius changes. The benchmark update strategy includes the gradient ascent method, moving the benchmark along the direction of increasing density, but this embodiment uses a random sampling method to avoid getting trapped in local optima. The entire dense scanning operation and positioning process is encapsulated into an independent software module. The module takes into account the current temperature stability value, humidity stability value, and time stability value, and outputs the optimal processing cycle. The module's execution time is optimized to the millisecond level to meet real-time processing requirements. A visualization tool for the parameter space is optionally provided, offering a graphical interface to display the benchmark movement trajectory and density distribution to assist in debugging and verification, but this is not a core function.

[0045] Example 4: Calculating Distribution Density. A spherical subspace is constructed with a reference point as its center. The reference point is determined in the three-dimensional parameter space by the current temperature stability, humidity stability, and time stability values. The construction of the spherical subspace relies on a preset search radius, which is a configurable parameter. A typical value for the preset search radius is set to 0.15 stability units. The volume of the spherical subspace is calculated using the formula for the volume of a sphere. Counting the number of basic data points contained within the spherical subspace requires calculating the Euclidean distance between each basic data point and the reference point. This distance calculation involves the square root of the sum of the squared differences in the three coordinate dimensions. Basic data points are counted when the distance between them and the reference point is less than or equal to the preset search radius. The counting process of basic data points is accelerated using a spatial index structure. The spatial index structure divides the three-dimensional parameter space into uniform grid cells. Each grid cell records a list of basic data points falling within it. During the count, only the basic data points within the grid cells covered by the spherical subspace need to be checked, avoiding traversing all basic data points. The local density value is calculated by dividing the number of basic data points by the volume of the spherical subspace. The local density value is a scalar value that characterizes the concentration of basic data points within a unit volume near the reference point. The calculation result of the local density value is the basis for subsequent iterative comparisons.

[0046] To determine the optimal processing cycle, an iterative optimization method is used to find the region with the highest local density. The iteration process starts from the current reference point, calculates the local density value at the current reference point, and records the local density value as the current optimal value. A basic data point is randomly selected as a candidate point from the boundary of the spherical search region. The boundary of the spherical search region is defined as a sphere with the reference point as the center and a preset search radius as the radius. The random selection of the candidate point is achieved through a spherical uniform sampling algorithm. The neighborhood density of the candidate point is calculated in the same way as the local density value. A new spherical subspace is constructed with the candidate point as the center. The radius of the new spherical subspace is the same as the radius of the original spherical search region. The number of basic data points in the new spherical subspace is counted and divided by the volume to obtain the neighborhood density. The neighborhood density of the candidate point is compared with the current local density value. If the neighborhood density is greater than or equal to the current local density value, the candidate point is updated as the new reference point, and the current local density value is updated with the neighborhood density. The iteration process is repeated. In each iteration, a new candidate point is randomly selected for comparison and possible updates. The number of iterations is limited by a maximum number of iterations, which is set to one hundred to prevent infinite loops. After the iteration terminates, the sample processing cycle corresponding to the final benchmark point is selected as the optimal processing cycle. The optimal processing cycle includes specific temperature control parameters, humidity control parameters, and processing time parameters.

[0047] The calculation of distribution density involves multiple parameters, the settings of which affect computational efficiency and result accuracy. The selection of the preset search radius needs to balance the search range and solution precision. A larger preset search radius includes more basic data points, and the calculated local density value reflects the average distribution over a larger area. A smaller preset search radius makes the local density value more focused on the distribution of points in neighboring regions. The volume calculation of the spherical subspace uses the sphere volume formula; the volume calculation is the denominator in the local density value calculation, and the accuracy of the volume value directly affects the numerical stability of the local density value. The statistical precision of the number of basic data points depends on the granularity of the spatial index structure. A spatial index structure with a smaller grid size provides more accurate statistical results but increases memory consumption and computation time. The calculation of local density values ​​uses double-precision floating-point arithmetic to avoid the accumulation of rounding errors. Local density values ​​are stored in temporary variables for iterative comparison. The implementation of the iterative optimization process requires maintaining multiple state variables, including the 3D coordinates of the current reference point, the current local density value, and the iteration counter. Candidate points are selected using a pseudo-random number generator. The generator produces uniformly distributed random direction vectors. These vectors are multiplied by a preset search radius to obtain a random offset on the sphere, which is then superimposed onto the baseline coordinates to obtain the candidate point coordinates. The neighborhood density calculation repeats the distribution density calculation process, but uses the candidate point as the new center. The neighborhood density calculation result is the basis for determining whether to update the baseline point. The baseline point update operation includes coordinate replacement and local density value update. After the baseline point is updated, the iteration counter is incremented, and the loop terminates when the iteration counter reaches the maximum number of iterations. The final baseline coordinates are used to query the nearest basic data point, and the sample processing cycle associated with the basic data point is extracted as the optimal processing cycle.

[0048] The process of determining the optimal processing cycle can use a parameter table to record intermediate results. The parameter table contains key parameters and calculation results during the iteration process. Refer to Table 1, which shows a parameter table recording parameter values ​​under different iteration steps.

[0049] Table 1: Parameters for Calculating Distribution Density

[0050] Iteration steps Reference point temperature coordinates Reference point humidity coordinates Reference point time coordinates Local density value Candidate point temperature coordinates Humidity coordinates of candidate points Candidate point time coordinates Neighborhood density Update? 1 0.50 0.60 0.55 12.5 0.52 0.58 0.53 13.2 yes 2 0.52 0.58 0.53 13.2 0.51 0.59 0.54 12.8 no 3 0.52 0.58 0.53 13.2 0.53 0.57 0.52 13.5 yes ... ... ... ... ... ... ... ... ... ... 100 0.56 0.54 0.50 15.3 0.55 0.55 0.51 15.0 no

[0051] The parameter table records the coordinates of the reference point, local density value, candidate point coordinates, neighborhood density, and update decision for each iteration. The parameter table is used to debug and verify the convergence of the iteration process. The reference point temperature coordinates, reference point humidity coordinates, and reference point time coordinates are the coordinate values ​​of the reference point in the 3D parameter space; the candidate point temperature coordinates, candidate point humidity coordinates, and candidate point time coordinates are the coordinate values ​​of the candidate point. The local density value is the distribution density at the reference point, the neighborhood density is the distribution density at the candidate point, and the update flag indicates whether a candidate point is adopted as a new reference point. The data in the parameter table is updated in real time during the iteration process, and the records in the parameter table help understand the movement trajectory and density change trend of the reference point. The calculation of distribution density and the determination of the optimal processing cycle rely on 3D geometric operations and iterative algorithms. 3D geometric operations include distance calculation between points, sphere volume calculation, and spherical sampling. Distance calculation uses the Euclidean distance formula, involving square and square root operations, and the operations use a hardware-accelerated mathematical function library. Sphere volume calculation is based on the sphere volume formula and is an important component of local density calculation. Spherical sampling generates uniformly distributed random points; the spherical sampling algorithm achieves this by generating random unit vectors. The convergence of the iterative algorithm depends on the density distribution characteristics of the parameter space. A smooth density distribution leads to rapid convergence, while a multi-peaked density distribution may get stuck in local optima. Setting a maximum number of iterations ensures the algorithm terminates within a finite time, and the output of the optimal processing cycle serves as a guiding parameter for optimizing the peptide content of bird's nest.

[0052] Example 5: The optimal processing cycle is derived from the output of a three-dimensional parameter space search, including a set of specific process parameters: the heating temperature is set at 62 degrees Celsius, the humidity level is maintained at 75% relative humidity, and the processing time is determined to be 360 ​​minutes. The bird's nest raw material uses dried white bird's nest from Southeast Asia. Before processing, the raw material is soaked in pure water for eight hours at a temperature maintained at 20 degrees Celsius. After soaking, the weight of the raw material increases to eight times its dry weight. The raw material is evenly placed on a specially designed stainless steel processing tray, with each tray measuring 40 cm x 60 cm and holding 100 grams of wet bird's nest. Peptide content optimization processing is carried out in a controlled, sealed processing chamber equipped with an infrared heating module, an ultrasonic humidification module, and a multi-point temperature and humidity sensor network. The infrared heating module has a power of 3 kilowatts and an adjustable heating range from room temperature to 100 degrees Celsius. The ultrasonic humidification module atomizes 1.5 liters per hour, with a humidity control accuracy of ±2%. At the start of processing, the environmental control unit adjusts the operating parameters of the processing chamber according to the optimal processing cycle. Heating temperature is achieved by adjusting the power supply voltage of the infrared heating tube, with the voltage value output by a digital-to-analog converter, corresponding to a set value of 62 degrees Celsius. Humidity is managed by the control circuit of the ultrasonic humidifier, which adjusts the transducer's operating frequency based on feedback signals from the humidity sensor, stabilizing the relative humidity inside the processing chamber at 75%. The processing time is precisely controlled by an industrial timer, which starts a 360-minute countdown at the start of processing, with a timing accuracy of ±0.5 seconds. Peptide extraction is performed in three consecutive batches. After each batch, the processing chamber automatically enters a five-minute cooling phase, during which heating is stopped and ventilation is maintained, allowing the bird's nest raw material temperature to naturally decrease to 35 degrees Celsius. After each batch, bird's nest samples are immediately taken from the tray, with sampling locations fixed at the four corners and the center of the tray. Three grams of sample are taken from each sampling point, and the samples are sealed in vacuum-sealed aluminum foil bags and marked with a timestamp.

[0053] The serialized processing dataset is dynamically generated during each processing round. This dataset includes time-series process parameter records and corresponding peptide concentration measurements. Process parameter records are acquired once per second, with each record including a timestamp, actual heating temperature, actual humidity value, heater power output percentage, and humidifier operating current. Peptide concentration data is acquired in real-time using an online near-infrared spectroscopy probe installed in the observation window of the processing chamber. The spectral acquisition wavenumber range is 1200 to 2500 per centimeter, with a resolution of 8 per centimeter. After each processing round, the laboratory performs offline validation analysis on the samples. Offline analysis uses high-performance liquid chromatography (HPLC) with the following chromatographic conditions: a C18 reversed-phase column, a column temperature of 40 degrees Celsius, a mobile phase of acetonitrile and 0.1% formic acid aqueous solution, and a gradient elution program running for 20 minutes. Online measurements are compared and calibrated with offline analysis values. The calibrated peptide concentration data is then incorporated into a serialized data set, which is stored in a structured database table format. Each record contains the processing round, time point, temperature, humidity, online peptide concentration, and calibrated peptide concentration.

[0054] After completing three rounds of processing, the peptide content determination unit initiates a comprehensive evaluation process. This unit consists of a spectral analysis module, a data statistics module, and a report generation module. The spectral analysis module performs point-by-point verification of the peptide concentration data in the serialized dataset. This verification includes signal quality assessment and outlier removal. Signal quality assessment checks the signal-to-noise ratio (SNR) of each spectral signal; signals with an SNR below 100:1 are marked as suspicious data. Outlier removal uses the Grubbs criterion, detecting and removing peptide concentration values ​​that deviate from the mean by more than three standard deviations. The verified data then enters the data statistics module, which calculates the average peptide content and deviation. The average peptide content is the arithmetic mean of all valid peptide concentration data, and the deviation is calculated using the sample standard deviation formula, with the degrees of freedom set to n minus one. Simultaneously, the data statistics module performs trend analysis, fitting a quadratic polynomial curve of peptide concentration over time and calculating the coefficient of determination to assess the goodness of fit. The report generation module integrates and analyzes the results to generate an assessment of the improvement in bird's nest peptide content. This assessment is output in a standardized report format, including header information, a data summary, statistical results, and quality indicators. The header information records the batch number, date and time, and operator ID. The data summary lists the total number of data points, the number of valid data points, the number of invalid data points, and the reasons for removal. The statistical results display the average peptide content, the deviation value, and the coefficient of determination of the trend equation. The quality indicators are graded based on the deviation value: a deviation value less than 0.05 is marked as excellent, a deviation value between 0.05 and 0.1 is marked as good, and a deviation value greater than 0.1 is marked as requiring improvement. The report is saved as a PDF file to the system log directory, and a paper copy is also printed for archiving.

[0055] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A method for assisting in increasing the peptide content of bird's nest, characterized in that, The method is implemented through the following process: The raw data set of bird's nest materials is collected by monitoring equipment within a preset historical period. The raw data set of the multi-parameter data includes historical temperature dataset, historical humidity dataset and historical time dataset. The overall stability of the multi-parameter raw data set is calculated, and the values ​​of temperature stability, humidity stability and time stability are derived. Using the temperature stability value, humidity stability value, and time stability value as guiding parameters, a dense scanning operation is performed within the predefined bird's nest processing parameter space to locate the optimal processing cycle. Based on the optimal processing cycle, the bird's nest raw materials are subjected to peptide content optimization treatment, a serialized processing data set is generated, and the peptide content determination unit is called to evaluate the serialized processing data set, and the evaluation result of the improvement of bird's nest peptide content is output.

2. The method for assisting in increasing the peptide content of bird's nest as described in claim 1, characterized in that, Acquiring a multi-parameter raw data set includes: deploying distributed sensing devices to continuously capture temperature change signals, humidity change signals, and time flow signals in the bird's nest processing environment, and archiving them into the historical temperature dataset, historical humidity dataset, and historical time dataset according to a preset sampling frequency.

3. The method for assisting in increasing the peptide content of bird's nest as described in claim 2, characterized in that, The overall stability calculation of the multi-parameter raw data set includes: calculating the average value and variation within the moving window for the historical temperature dataset, historical humidity dataset, and historical time dataset respectively, and merging the average value and variation to generate the temperature stability value, humidity stability value, and time stability value.

4. The method for assisting in increasing the peptide content of bird's nest as described in claim 3, characterized in that, Calculating the average and variation within a moving window includes: setting a fixed-length sliding time window, and calculating the average temperature and temperature variation of the historical temperature dataset, the average humidity and humidity variation of the historical humidity dataset, and the average time and time variation of the historical time dataset window by window.

5. The method for assisting in increasing the peptide content of bird's nest as described in claim 1, characterized in that, Performing intensive scanning operations within a predefined bird's nest processing parameter space includes: establishing a three-dimensional parameter space with temperature stability values ​​as the horizontal axis, humidity stability values ​​as the vertical axis, and time stability values ​​as the vertical axis; and loading multiple sample stability values ​​and corresponding sample processing cycles as basic data points in the three-dimensional parameter space.

6. The method for assisting in increasing the peptide content of bird's nest as described in claim 5, characterized in that, Determining the optimal processing cycle includes: using the current temperature stability value, humidity stability value, and time stability value as reference points, defining a spherical search region in the three-dimensional parameter space, and calculating the distribution density of basic data points within the spherical search region.

7. The method for assisting in increasing the peptide content of bird's nest as described in claim 6, characterized in that, The calculation of distribution density includes: constructing a spherical subspace with the reference point as the center and a preset search radius as the scale; counting the number of basic data points contained in the spherical subspace, and dividing the number of basic data points by the volume of the spherical subspace to obtain the local density value.

8. The method for assisting in increasing the peptide content of bird's nest as described in claim 7, characterized in that, The optimal processing cycle for locating the target point also includes: randomly selecting a basic data point as a candidate point from the boundary of the spherical search region, calculating the neighborhood density of the candidate point; if the neighborhood density is greater than or equal to the current local density value, updating the candidate point as a new reference point, repeating the density comparison process until the maximum number of iterations is reached, and taking the sample processing cycle corresponding to the final reference point as the optimal processing cycle.

9. The method for assisting in increasing the peptide content of bird's nest as described in claim 1, characterized in that, The process of optimizing the peptide content of bird's nest raw materials according to the optimal processing cycle includes: adjusting the heating temperature, moisturizing level and processing time of bird's nest according to the optimal processing cycle, performing multiple rounds of peptide extraction operations, and recording the peptide concentration data after each round of operations to form the serialized processing data set.

10. The method for assisting in increasing the peptide content of bird's nest as described in claim 9, characterized in that, The evaluation process using the peptide content determination unit includes: employing spectral analysis technology to verify the peptide concentration data in the sequenced data set point by point, calculating the average peptide content and deviation value, and generating the evaluation result of the improved bird's nest peptide content based on the average peptide content and deviation value.

Images