Intelligent control system for water environment monitoring

Abstract

The invention relates to the technical field of water quality monitoring, and discloses an intelligent control system for water environment monitoring, which is characterized in that an upstream monitoring unit and a downstream monitoring unit are used for acquiring a salinity time sequence, a signal processing module is used for filtering data to extract a subtidal component, and a consciousness reference phase is obtained through Hilbert transformation; calculating a phase difference and converting the phase difference into an instant travel delay function; the fusion processing module carries out time axis mapping and persistence weighting based on a time delay function to obtain a fusion salinity time sequence; and finally, the intelligent control module implements discharge regulation and control according to a fusion result, so that dynamic alignment and precise fusion of upstream and downstream salinity signals are realized, and the precision and reliability of monitoring and water environment regulation and control are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of water quality monitoring, and more particularly to an intelligent control system for water environment monitoring. Background Art

[0002] With the advancement of comprehensive river basin management and water ecological restoration, multi-node sensor monitoring systems have been widely deployed in tidal estuaries and downstream mainstream areas, which serve as key nodes for water environment regulation. These systems can collect core water quality indicators such as salinity, temperature, dissolved oxygen, and nutrients in real time. Their goal is to use multi-dimensional data to analyze pollutant transport patterns and predict water quality trends, providing basic data support for river basin water environment management. Conventional thinking often considers water quality signals from upstream monitoring nodes as the input source of downstream signals, with only a simple time transmission delay between the two. Therefore, traditional monitoring solutions often design data processing logic based on this.

[0003] However, the unique hydrodynamic environment of tidal river sections presents significant complexity in water environment monitoring. On the one hand, tidal movement causes continuous fluctuations in downstream hydrodynamic conditions, creating a dynamic game between the freshwater dynamics of the upstream inflow and the saltwater dynamics of the downstream tidal current, disrupting the stable upstream-downstream transmission relationship. On the other hand, the saltwater intrusion process exhibits distinct asymmetric characteristics within the semi-lunar tidal cycle. Specifically, salinity rises rapidly during the transition from neap to spring tide, while the rate of salinity decreases significantly during the transition from spring to neap tide. This dynamic process of rapid rise and slow fall causes upstream and downstream water quality signals on long timescales to no longer maintain a simple lag relationship. Instead, they experience a reversal of dependence and a continuous spatial and temporal misalignment.

[0004] Current mainstream water environment monitoring methods suffer from two major limitations: First, the use of fixed time lags to align upstream and downstream monitoring data ignores the dynamic changes in hydrodynamics and salinity over a semi-lunar period in tidal river sections, resulting in insufficient data alignment accuracy. Second, the direct use of simultaneous data for cross-section comparisons fails to reflect the true response relationship between upstream and downstream water quality signals. Both of these methods are prone to significant errors in the context of semi-lunar-scale tidal and salinity variations. They can misinterpret upstream salinity anomalies as independently generated downstream, and can also misproject the arrival times of pollution peaks or salt fronts, distorting basin water quality load calculations. These methods fail to meet the practical requirements for intelligent control for precise water environment management in tidal river sections. This can lead to uncontrolled salinity in tidal sections, further causing technical problems related to aquatic environmental imbalance in these sections. Summary of the Invention

[0005] The present invention provides an intelligent control system for water environment monitoring, which solves the technical problems raised in the background technology.

[0006] The present invention provides an intelligent control system for water environment monitoring, comprising: The upstream monitoring unit is located at the upstream section near the tidal boundary and collects upstream salinity time series; The downstream monitoring unit is located at the downstream section near the tidal boundary and collects the downstream salinity time series; The signal processing module is connected to the upstream monitoring unit and the downstream monitoring unit and is configured to: The upstream and downstream salinity time series are filtered to separate the subtidal components. Based on the subtidal components, the upstream and downstream syzygy reference phases are obtained through Hilbert transform. The cross-node and cross-layer syzygy phase differences are calculated based on the upstream and downstream syzygy reference phases, and the syzygy phase differences are converted into instantaneous travel time lag functions based on the syzygy beat frequency. The fusion processing module is connected to the signal processing module and is configured as follows: Taking the synodic period as the time window, the persistence parameter is calculated according to the symbol duration of the immediate travel time lag function. The upstream salinity time series is time-axis mapped based on the immediate travel time lag function to obtain the aligned upstream salinity time series. The fusion weight coefficient is determined according to the persistence parameter, and the aligned upstream salinity time series is fused with the downstream salinity time series according to the fusion weight coefficient to obtain the fused salinity time series. Intelligent control module that performs discharge processing based on the fused salinity time series.

[0007] Furthermore, the upstream salinity time series and the downstream salinity time series are filtered to separate the subtidal components, including: Based on the astronomical angular frequency set Determine the upper limit of the low-pass passband and the start of the stopband ;in, , represents the angular frequency of the M2 tidal component, represents the angular frequency of the S2 tidal component, represents the angular frequency of the K1 tidal component, represents the angular frequency of the O1 partial tide; Perform fast Fourier transform on the upstream salinity time series and the downstream salinity time series to obtain the upstream frequency domain spectrum and downstream frequency domain spectrum ; Constructing a zero-phase low-pass transfer function ;in, exist When the value is 1, The value is 0 when When the cosine smooth transition is obtained, the upstream low-pass main filter is calculated. and downstream low-pass main filter ;in, Belongs to the upstream frequency domain spectrum or the downstream frequency domain spectrum; right and Applying a narrowband notch transfer function ,as follows: in, Indicates the M2 tidal or S2 tidal notch bandwidth; The product of the narrowband notch transfer function and the zero-phase low-pass transfer function is taken as the total transfer function; The upstream subtidal component is obtained by performing an inverse fast Fourier transform on the product of the total transfer function and the upstream low-pass main filter; the downstream subtidal component is obtained by performing an inverse fast Fourier transform on the product of the total transfer function and the downstream low-pass main filter.

[0008] Furthermore, based on the subtidal component, the upstream syzygy reference phase and the downstream syzygy reference phase are obtained by Hilbert transform, including: The upstream subtidal component and the downstream subtidal component are removed from the mean and standardized to obtain the standard upstream subtidal component and standard downstream subtidal component ; Will and The absolute difference is taken as the syzygy beat frequency; Generates a time-varying reference waveform using the sine value of the syzygy beat frequency , and calculate the reference phase of the reference waveform through Hilbert transform ; The upstream analytical signals of the standard upstream subtidal component and the standard downstream subtidal component are constructed using the reference phase. and downstream analytical signals ;in, represents the Hilbert transform operation, i is the imaginary unit, ; Calculate the upstream instantaneous phase of the upstream analytical signal , calculate the downstream instantaneous phase of the downstream analytical signal ;in, represents the binary inverse tangent function, represents the imaginary part of the analytical signal, Real part of the analytical signal; Calculate upstream syzygy reference phase ; Calculate downstream syzygy reference phase .

[0009] Furthermore, the cross-node and cross-layer syzygy phase difference is calculated based on the upstream syzygy reference phase and the downstream syzygy reference phase, and the syzygy phase difference is converted into an instantaneous travel time delay function based on the syzygy beat frequency, including: The phase difference between the upstream syzygy reference phase and the downstream reference phase is obtained by the wrap function to obtain the main value phase difference; wherein, the wrap function maps the phase difference value to the interval ; The main value phase difference is continuously expanded by unwrap to obtain the continuous phase difference; The ratio of the continuous phase difference to the syzygy beat frequency is used as the instantaneous time delay function .

[0010] Furthermore, taking the synodic period as the time window, the duration parameter is calculated according to the symbol duration of the instantaneous travel time lag function, including: The ratio of 2π to the synodic beat frequency is taken as the synodic period ; By time Defining a time window ; Through the time window Collect the instantaneous time-lag function value at fixed time intervals; If the value of the instantaneous time-history lag function is greater than 0, the sign is positive 1; If the value of the instantaneous time-lag function is less than 0, the sign is negative 1; If the value of the instantaneous time-lag function is equal to 0, the difference between the two adjacent instantaneous time-lag function values ​​on the time scale is calculated, and the corresponding sign is determined according to the positive or negative value of the difference; Determine the moment when the value of the instantaneous time-history lag function is positive 1 to obtain a first duration; determining a moment when the value of the instantaneous time-history lag function has a sign of negative 1 to obtain a second duration; The ratio of the difference between the first duration and the second duration to the synodic period is taken as the moment The persistence parameter.

[0011] Furthermore, the upstream salinity time series is mapped to the time axis based on the instantaneous travel time lag function to obtain the aligned upstream salinity time series, including: Calculate the difference between the upstream salinity value at the sth moment in the upstream salinity time series and the instantaneous time history lag function value at the sth moment, so as to form an initial salinity alignment sequence based on the difference; If the initial salinity alignment value at the sth moment in the initial alignment sequence is ≥ the initial salinity alignment value at the s+1th moment, the initial salinity alignment sequence is optimized, including: The first constraint includes: the initial salinity alignment value at the sth moment < the initial salinity alignment value at the s+1th moment; The second constraint includes: calculating the square value of the initial salinity alignment value at the s-th moment after optimization and the initial salinity alignment value at the s-th moment, and calculating the sum of the square values ​​at each moment; and minimizing the sum of the square values; An optimized salinity alignment sequence is formed based on the first constraint and the second constraint to obtain an aligned upstream salinity time series.

[0012] Furthermore, the fusion weight coefficient is determined according to the persistence parameter, and the aligned upstream salinity time series and the downstream salinity time series are fused according to the fusion weight coefficient to obtain the fused salinity time series, including: Construct a time window at the sth moment to determine the duration parameter at the sth moment ; Construct the fusion weight coefficient with the duration parameter at the sth moment ; and limit the maximum value of the fusion weight coefficient to 1 and the minimum value to 0; Calculate the aligned upstream salinity value at time s and The first product of and the downstream salinity value at the sth moment and The second product of the first and second products is taken as the fused salinity value at the sth moment to form a fused salinity time series.

[0013] Furthermore, discharge processing is performed based on the fused salinity time series, including: Determine the fused salinity value corresponding to the fused salinity time series at the current moment; If the fused salinity value is not within the preset salinity range, the flow of the upstream outflow port is controlled accordingly until the fused salinity value is within the preset salinity range.

[0014] The beneficial effects of the present invention are: by constructing the synodic phase difference, the instantaneous travel time lag function, the persistence parameter and the dynamic weight fusion, the dynamic phase alignment and weight distribution of the salinity signals upstream and downstream of the tidal river section are realized, so as to accurately capture the asymmetric response properties of upstream leading and downstream lagging under the drive of the semi-lunar tide, avoid dislocation and systematic deviation, significantly improve the accuracy of monitoring data alignment and the reliability of the fusion results, and can intelligently control the discharge based on this, so that the salinity is kept within a reasonable range, effectively supporting the precise regulation of the water environment and ecological stability of the tidal section. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Figure 1 This is a module diagram of an intelligent control system for water environment monitoring according to the present invention. DETAILED DESCRIPTION

[0016] The subject matter described herein will now be discussed with reference to exemplary embodiments. It should be understood that these embodiments are discussed solely to enable those skilled in the art to better understand and implement the subject matter described herein, and that the functions and arrangements of the elements discussed may be varied without departing from the scope of this specification. Various examples may omit, substitute, or add various processes or components as needed. In addition, features described with respect to some examples may also be combined in other examples.

[0017] like Figure 1 As shown, an intelligent control system for water environment monitoring includes: The upstream monitoring unit is located at the upstream section near the tidal boundary and collects upstream salinity time series; The downstream monitoring unit is located at the downstream section near the tidal boundary and collects the downstream salinity time series; The signal processing module is connected to the upstream monitoring unit and the downstream monitoring unit and is configured to: The upstream and downstream salinity time series are filtered to separate the subtidal components. Based on the subtidal components, the upstream and downstream syzygy reference phases are obtained through Hilbert transform. The cross-node and cross-layer syzygy phase differences are calculated based on the upstream and downstream syzygy reference phases, and the syzygy phase differences are converted into instantaneous travel time lag functions based on the syzygy beat frequency. The fusion processing module is connected to the signal processing module and is configured as follows: Taking the synodic period as the time window, the persistence parameter is calculated according to the symbol duration of the immediate travel time lag function. The upstream salinity time series is time-axis mapped based on the immediate travel time lag function to obtain the aligned upstream salinity time series. The fusion weight coefficient is determined according to the persistence parameter, and the aligned upstream salinity time series is fused with the downstream salinity time series according to the fusion weight coefficient to obtain the fused salinity time series. Intelligent control module that performs discharge processing based on the fused salinity time series.

[0018] It should be noted that the persistence of semi-lunar memory phase reversal is a phenomenon in the salinity dynamics of tidal river sections. Specifically, the phase change of the upper bottom salinity within the semi-lunar period leads that of the lower middle and lower layers downstream for a long time, and the sign of this leading and lagging relationship can be maintained across multiple tidal cycles. Its physical mechanisms include: the asymmetry and hysteresis effect of salt intrusion under the modulation of the spring tide; the alternation of tide-driven and density-driven mechanisms during strong and weak tides, the response amplification caused by the reflection of tidal waves in the upper section, and the residual transport generated by the lateral salt storage and discharge process on the tidal flat. The combined effect of these factors makes the phase reversal state significantly persistent, resulting in physical dislocation and systematic deviation in the traditional multi-node fusion method based on fixed time lag or synchronous time, and a new monitoring strategy with dynamic phase alignment and persistence weighting is needed.

[0019] It is important to note that the upstream section near the tidal boundary, located at the upstream end of the tidal reach, is a monitoring section near the extreme boundary of tidal influence propagating upward along the river channel and serves as the transition zone between runoff and tidal dynamics. The upstream section near the tidal boundary is characterized by runoff dominance and tidal disturbance. Upstream runoff (freshwater) still holds a certain preponderance, and the overall flow trend is downstream. However, during high tide, tidal uplift may cause brief landward (upstream) backflow at the bottom layer (especially during neap tides, when density driving is enhanced), forming the rudiments of a vertical circulation with the surface layer moving downward and the bottom layer moving upward. While salinity at the upstream section near the tidal boundary is generally low (due to dilution by upstream freshwater), bottom salinity is more sensitive to saltwater intrusion (saltwater, due to its high density, preferentially propagates along the bottom layer), resulting in a predisposing signal of a leading rise from neap to spring tide. This serves as the initial source of the upstream leading characteristic of the persistent semi-lunar memory phase reversal phenomenon.

[0020] It should be noted that the downstream section near the tidal boundary is located at the downstream end of the tidal river section, close to the estuary or the core area of ​​tidal action. It is a monitoring section in a typical response area where tidal dynamics dominate and runoff influence is relatively weak. The downstream section near the tidal boundary is characterized by tide dominance and runoff modulation. The reciprocating movement of the tide is violent, and the flow velocity and direction are periodically reversed (landward at high tide and seaward at low tide). The tidal range is large, the tidal velocity is high, and the tidal dispersion effect is significant. The salinity of the downstream section near the tidal boundary is generally high (affected by salt water from the open sea). The salinity in the middle and lower layers is affected by the density circulation (landward) and tidal mixing (vertical), and the changes lag behind the upstream bottom layer. In addition, due to river channel propagation and tidal modulation, the salinity phase will show delayed accumulation.

[0021] Specifically, the calibration steps for the upstream section near the tidal boundary are as follows: Initial scope locking: temporary monitoring points (intervals of 500m to 1km) are set up along the river channel from upstream to downstream (towards the estuary), and continuous monitoring is carried out for 7 complete tidal cycles (covering at least 1 synodic cycle), recording the flow velocity (vertical stratification), salinity (vertical stratification) and water level data at each point.

[0022] Calculate the proportion of landward flow velocity in the bottom layer during high tide at each point: the ratio of the duration of landward flow velocity to the total time of high tide. If the ratio is ≤30%, it means that the tidal support is only a short-term disturbance and runoff is still dominant.

[0023] Calculate the ratio of tidal amplitude to runoff velocity: the ratio of the water level amplitude caused by the tide to the average runoff velocity of the section. If the ratio is ≤0.5 (indicating that the runoff dynamics are still stronger than the tidal disturbance),

[0024] The maximum bottom salinity is ≤5‰ (significantly diluted by upstream freshwater), and the bottom salinity rise rate from neap tide to spring tide is more than 20% faster than that of the middle layer (reflecting the advanced response of the bottom layer to salt intrusion).

[0025] The most downstream point that meets the above-mentioned dynamic and salinity thresholds is the upstream tidal boundary section.

[0026] Specifically, the calibration steps for the downstream section near the tidal boundary are as follows: Initial scope locking: temporary monitoring points (at intervals of 500m to 1km) were set up along the river from downstream to upstream (away from the estuary), and continuous monitoring was also carried out for 7 complete tidal cycles.

[0027] Calculate the proportion of landward flow velocity in the bottom layer during high tide at each point: the ratio of the duration of the landward flow velocity to the total time of high tide. If the ratio is ≥60%, it indicates that the tidal support is significant and the landward flow dominates the high tide period.

[0028] Calculate the tidal amplitude to runoff velocity ratio: the ratio of the water level amplitude caused by the tide to the average runoff velocity of the section. If this value is ≥ 2.0 (indicating that tidal dynamics dominate),

[0029] The maximum salinity of the middle and lower layers is ≥10% (significantly affected by salt water from the open sea), and the rising rate of salinity in the middle and lower layers from neap tide to spring tide lags behind the salinity of the upstream bottom layer (the lag time is ≥12 hours, reflecting a phase delay).

[0030] The most upstream point that meets the above-mentioned dynamic and salinity thresholds is the downstream tidal boundary section.

[0031] In one embodiment of the present invention, filtering is performed on the upstream salinity time series and the downstream salinity time series to separate the subtidal components, including: Based on the astronomical angular frequency set Determine the upper limit of the low-pass passband and the start of the stopband ;in, , represents the angular frequency of the M2 tidal component, represents the angular frequency of the S2 tidal component, represents the angular frequency of the K1 tidal component, represents the angular frequency of the O1 partial tide; In detail, the M2 tide is the main semi-diurnal tide produced by the lunar tidal force, with a period of about 12 hours and 25 minutes. It is one of the most significant tidal components in the ocean and tidal river sections, and its angular frequency is about rad / s. The S2 tide is the main semi-diurnal tide produced by the solar tidal force. Its period is close to that of the M2 tide, about 12 hours. Its angular frequency is about rad / s, is the most important semi-diurnal tidal component after the M2 tide. The K1 tide is the main diurnal tide produced by the combined effects of the lunar and solar tidal forces, with a period of about 23 hours and 56 minutes, and its angular frequency is about rad / s, which is a kind of high energy component of the diurnal tide. The O1 tide is the diurnal tide produced by the declination component of the lunar tidal force, with a period of about 25 hours and 49 minutes, and its angular frequency is about rad / s, which is also a common diurnal tidal component in tidal waters.

[0032] It should be noted that the upper limit of the low-pass passband must be less than the starting point of the stopband, and both must be less than the smallest angular frequency in the set. This strictly limits the passband to below the diurnal and semi-diurnal tide frequencies, ensuring that only slowly varying signals with periods longer than these high-frequency tides, namely subtidal components, are retained; the stopband covers all major diurnal and semi-diurnal tide frequencies, thus suppressing high-frequency interference.

[0033] Perform fast Fourier transform on the upstream salinity time series and the downstream salinity time series to obtain the upstream frequency domain spectrum and downstream frequency domain spectrum ; It should be noted that the upstream and downstream salinity time series are subjected to fast Fourier transforms to obtain the upstream and downstream frequency domain spectra. The time domain salinity series contains signal components of different frequencies, and the fast Fourier transform can decompose it into a frequency domain spectrum, visually displaying the energy distribution of each frequency component.

[0034] Constructing a zero-phase low-pass transfer function ;in, exist When the value is 1, The value is 0 when When the cosine smooth transition is obtained, the upstream low-pass main filter is calculated. and downstream low-pass main filter ;in, Belongs to the upstream frequency domain spectrum or the downstream frequency domain spectrum; In detail, a zero-phase low-pass transfer function is constructed. The zero-phase low-pass transfer function takes the value of 1 when the absolute value of the frequency is less than or equal to the upper limit of the low-pass passband, indicating that the low-frequency signals within the range are completely retained; the zero-phase low-pass transfer function takes the value of 0 when the absolute value of the frequency is greater than or equal to the start point of the stopband, indicating that the high-frequency signals within the range are completely filtered out; between the upper limit of the low-pass passband and the start point of the stopband, a cosine smooth transition is used to reduce the oscillation distortion caused by frequency band mutation during the filtering process.

[0035] Based on the zero-phase low-pass transfer function, the upstream and downstream low-pass main filters are calculated, where represents the frequency in the upstream or downstream frequency domain spectrum. This low-pass filtering preliminarily suppresses high-frequency components such as diurnal and semi-diurnal tides. At the same time, the zero-phase characteristic ensures that the temporal relationship of the time series is not distorted, ensuring that the relative temporal relationship between the upstream and downstream signals is not affected by filtering.

[0036] right and Applying a narrowband notch transfer function ,as follows: in, Indicates the M2 tidal or S2 tidal notch bandwidth; In detail, the narrow-band notch transfer function is applied to the angular frequency of the M2 and S2 tidal components, and its expression is: ;in This represents the notch bandwidth for the M2 or S2 subtidal components and must be sufficiently narrow to avoid affecting signals within the lowpassband. The narrowband notch transfer function is used to suppress spectral leakage in the frequency domain from the M2 and S2 subtidal components (the strongest semidiurnal tidal components). Even after lowpass filtering, these strong high-frequency components may still retain residual energy in the transition region between the passband and stopband. The narrowband notch can specifically attenuate this residual interference, improving the purity of the subtidal components.

[0037] The product of the narrowband notch transfer function and the zero-phase low-pass transfer function is taken as the total transfer function; Specifically, the narrowband notch transfer function and the zero-phase low-pass transfer function are multiplied to obtain the total transfer function. The total transfer function combines the overall suppression of the high-frequency range by the low-pass filter with the precise attenuation of specific strong high-frequency components by the narrowband notch, achieving a synergistic filtering effect of broad-spectrum suppression and precise removal, minimizing the interference of high-frequency tidal signals on subtidal components.

[0038] The upstream subtidal component is obtained by performing an inverse fast Fourier transform on the product of the total transfer function and the upstream low-pass main filter; the downstream subtidal component is obtained by performing an inverse fast Fourier transform on the product of the total transfer function and the downstream low-pass main filter.

[0039] In detail, the inverse fast Fourier transform converts the processed signal in the frequency domain back to the time domain, and the final output subtidal components of the upstream and downstream subtidal components only contain slowly varying information with periods longer than the diurnal and semidiurnal tides.

[0040] In one embodiment of the present invention, obtaining an upstream syzygy reference phase and a downstream syzygy reference phase by Hilbert transform based on a subtidal component includes: The upstream subtidal component and the downstream subtidal component are removed from the mean and standardized to obtain the standard upstream subtidal component and standard downstream subtidal component ; It should be noted that the upstream and downstream subtidal components are demeaned and standardized to eliminate the interference of amplitude differences on phase analysis. Demeaning refers to subtracting the mean value from the subtidal component series to remove the overall offset; standardization refers to dividing the demeaned series by its standard deviation to bring the amplitudes of different series to the same order of magnitude. The standard upstream and downstream subtidal components obtained after demeaning and standardization retain only the phase characteristics and shape information of the signals, preventing amplitude differences caused by differences in monitoring site environments or instrument sensitivity from affecting subsequent phase comparisons.

[0041] Will and The absolute difference is taken as the syzygy beat frequency; Specifically, the synodic beat frequency is determined by the absolute difference between the angular frequencies of the M2 and S2 components. The M2 and S2 components are the two most significant semidiurnal tidal components, and the beat frequency resulting from this difference in angular frequency corresponds to a semilunar period (approximately 14.77 days). This semilunar period is a key indicator of the duration of the semilunar memory-type phase reversal phenomenon in tidal river salinity variations. Using this semilunar period as the synodic beat frequency ensures the stability and traceability of the reference standard.

[0042] Generates a time-varying reference waveform using the sine value of the syzygy beat frequency , and calculate the reference phase of the reference waveform through Hilbert transform ; Specifically, a sinusoidal reference waveform is generated with the synodic beat frequency as the angular frequency. The reference waveform has a strict semi-monthly period and a clear phase variation pattern, which can reflect the phase evolution of the synodic rhythm under ideal conditions. The reference waveform is processed through the Hilbert transform to obtain an analytical signal of the reference waveform, and then the reference phase is calculated based on the analytical signal. The reference phase is a physical quantity that describes the change of the phase of the reference waveform over time. It is used to reflect the angular change of the reference waveform in the complex plane, providing a reference benchmark phase scale for the phase of the upstream and downstream subtidal components, eliminating the phase incomparability problem caused by the difference in the starting time of different sequences.

[0043] The upstream analytical signals of the standard upstream subtidal component and the standard downstream subtidal component are constructed using the reference phase. and downstream analytical signals ;in, represents the Hilbert transform operation, i is the imaginary unit, ; Specifically, the analytical signal is a complex signal that describes the phase and amplitude characteristics of a real signal and consists of a real part and an imaginary part. For the standard upstream subtidal component, the real part of the analytical signal is the standard upstream subtidal component itself, and the imaginary part is the result of the Hilbert transform of this standard upstream subtidal component. Similarly, the real part of the analytical signal for the standard downstream subtidal component is the standard downstream subtidal component, and the imaginary part is the result of its Hilbert transform. The purpose of the Hilbert transform is to generate a component orthogonal to the original signal, so that the analytical signal can fully characterize the phase variation trajectory of the original signal in the complex plane, thereby accurately extracting the instantaneous phase.

[0044] Calculate the upstream instantaneous phase of the upstream analytical signal , calculate the downstream instantaneous phase of the downstream analytical signal ;in, represents the binary inverse tangent function, represents the imaginary part of the analytical signal, Real part of the analytical signal; Specifically, the instantaneous phase represents the phase state of the signal at a specific moment. It is calculated by calculating the phase of the analytical signal using the binary inverse tangent function. The binary inverse tangent function uses the imaginary part of the analytical signal as the numerator and the real part as the denominator. The result reflects the angle corresponding to the analytical signal in the complex plane, that is, the phase of the signal at that moment. For the upstream analytical signal, the calculated upstream instantaneous phase describes the phase change of the standard upstream subtidal component over time; the downstream instantaneous phase describes the phase change of the standard downstream subtidal component. This converts the complex plane characteristics of the analytical signal into a quantifiable phase value.

[0045] Calculate upstream syzygy reference phase ; Specifically, the instantaneous phase represents the phase state of the signal at a specific moment. It is calculated by calculating the phase of the analytical signal using the binary inverse tangent function. The binary inverse tangent function uses the imaginary part of the analytical signal as the numerator and the real part as the denominator. The result reflects the angle corresponding to the analytical signal in the complex plane, that is, the phase of the signal at that moment. For the upstream analytical signal, the calculated upstream instantaneous phase describes the phase change of the standard upstream subtidal component over time; the downstream instantaneous phase describes the phase change of the standard downstream subtidal component. This converts the complex plane characteristics of the analytical signal into a quantifiable phase value.

[0046] Calculate downstream syzygy reference phase .

[0047] Specifically, the upstream synodic reference phase is obtained by subtracting the reference phase from the upstream instantaneous phase, modulo 2π; the downstream synodic reference phase is obtained by subtracting the reference phase from the downstream instantaneous phase, modulo 2π. This unifies the upstream and downstream instantaneous phases to the reference phase, eliminating absolute phase differences and retaining only the phase offset relative to the synodic rhythm. Modulo 2π is used to constrain the phase values ​​to the range of 0 to 2π, ensuring phase periodicity and comparability. The resulting upstream and downstream synodic reference phases directly reflect the phase lead or lag relationship of the upstream and downstream subtidal components relative to the semilunar rhythm.

[0048] In one embodiment of the present invention, a cross-node and cross-layer syzygy phase difference is calculated based on the upstream syzygy reference phase and the downstream syzygy reference phase, and the syzygy phase difference is converted into an instantaneous travel time delay function based on the syzygy beat frequency, including: The phase difference between the upstream syzygy reference phase and the downstream reference phase is obtained by the wrap function to obtain the main value phase difference; wherein, the wrap function maps the phase difference value to the interval ; It should be noted that although both the upstream syzygy reference phase and the downstream syzygy reference phase describe the phase state with the reference phase of the reference waveform, directly calculating the difference between the two may produce results outside the normal range due to the periodicity of the phase. The phase is periodic and its value is usually constrained to be within a specific range. Direct subtraction may produce unreasonable differences, such as a sudden jump from close to π to close to -π. The wrap function is used to map this direct difference to By adjusting the difference to conform to the inherent laws of phase periodicity, the phase difference is ensured to be within a uniform and reasonable range. This eliminates the ambiguity of the difference caused by phase periodicity and puts the phase difference between upstream and downstream on a directly comparable scale.

[0049] The main value phase difference is continuously expanded by unwrap to obtain the continuous phase difference; Specifically, although the main value phase difference is already within a unified interval, the difference between adjacent moments may experience jumps due to crossing the interval boundary. For example, the main value phase difference between adjacent moments may suddenly change from near π to near -π. This jump is not a true phase change, but rather caused by the interval constraint. The unwrap function detects jumps in the main value phase difference between adjacent moments and automatically adds or subtracts 2π to eliminate the jump when the absolute value of the jump exceeds π. The unwrap function converts the originally discrete main value phase difference into a continuous phase difference that changes continuously over time. This restores the true change trend of the phase difference over time and ensures that the phase difference reflects the continuous evolution of the upstream and downstream phase relationship.

[0050] The ratio of the continuous phase difference to the syzygy beat frequency is used as the instantaneous time delay function .

[0051] Specifically, continuous phase differences are essentially angular differences that must be converted into time quantities to directly describe the lead or lag relationship between upstream and downstream. The synodic beat frequency is an angular frequency (measured in radians per second) based on astronomical constants. Its relationship to the period is that the angular frequency equals 2π divided by the period. Based on the physical relationship between phase and time, the phase difference equals the angular frequency multiplied by the time difference, and thus the time difference equals the phase difference divided by the angular frequency. Dividing the continuous phase difference by the synodic beat frequency yields the instantaneous travel lag function. The instantaneous travel lag function provides a real-time reflection of the upstream's lead or lag relative to the downstream's.

[0052] In one embodiment of the present invention, the duration parameter is calculated based on the symbol duration of the immediate travel time lag function using the synodic period as a time window, including: The ratio of 2π to the synodic beat frequency is taken as the synodic period ; Specifically, the conversion relationship between angular frequency and synodic period is: the synodic period equals 2π divided by the angular frequency. Therefore, the ratio of 2π to the synodic beat frequency is defined as the synodic period. The synodic period is determined by astronomical constants and reflects the inherent time scale of the semilunar rhythm.

[0053] By time Defining a time window ; Specifically, starting from any moment, a time window is defined as the interval between that moment and the synodic period. The length of the time window is strictly equal to the synodic period, ensuring that the time lag characteristics of the complete semilunar rhythm are counted within each window.

[0054] Through the time window Collect the instantaneous time-lag function value at fixed time intervals; If the value of the instantaneous time-history lag function is greater than 0, the sign is positive 1; If the value of the instantaneous time-lag function is less than 0, the sign is negative 1; If the value of the instantaneous time-lag function is equal to 0, the difference between the two adjacent instantaneous time-lag function values ​​on the time scale is calculated, and the corresponding sign is determined according to the positive or negative value of the difference; Specifically, the sign is used to distinguish between positive and negative trends in the instantaneous time lag function. When the value of the instantaneous time lag function is greater than 0, the sign is positive 1, indicating that the upstream is leading the downstream. When the value is less than 0, the sign is negative 1, indicating that the upstream is lagging the downstream. When the value is equal to 0, the sign is determined by calculating the difference between two adjacent instantaneous time lag function values ​​at that moment in time.

[0055] For example, for the s-1th moment, the sth moment and the s+1th moment, the difference between the instantaneous time lag function value at the s+1th moment and the instantaneous time lag function value at the s-1th moment is calculated; if the difference is a positive integer, the sign is positive 1, otherwise the sign is negative 1.

[0056] Determine the moment when the value of the instantaneous time-history lag function is positive 1 to obtain a first duration; determining a moment when the value of the instantaneous time-history lag function has a sign of negative 1 to obtain a second duration; Specifically, the first duration is the total time corresponding to all moments in the time window where the sign of the instantaneous time-lag function is positive 1. This is calculated by counting the number of sampling points in the window with a sign of positive 1 and multiplying it by a fixed time interval. The second duration is the total time corresponding to all moments in the time window where the sign is negative 1. This is also calculated by counting the number of sampling points with a sign of negative 1 and multiplying it by a fixed time interval. The first duration and the second duration respectively represent the actual durations of the upstream leading and lagging states within the window.

[0057] The ratio of the difference between the first duration and the second duration to the synodic period is taken as the moment The persistence parameter.

[0058] Specifically, the duration parameter is obtained by dividing the difference between the first duration and the second duration by the synodic period. This normalizes the duration difference to When the persistence parameter is greater than 0, it indicates that the upstream is generally in a leading state within that window; when it is less than 0, it indicates that the upstream is generally in a lagging state. The closer the absolute value is to 1, the stronger the persistence of the corresponding leading or lagging state. The persistence parameter converts the persistence characteristics of the time lag symbol into a quantifiable and comparable indicator.

[0059] In one embodiment of the present invention, the upstream salinity time series is mapped to a time axis based on the instantaneous travel lag function to obtain an aligned upstream salinity time series, including: Calculate the difference between the upstream salinity value at the sth moment in the upstream salinity time series and the instantaneous time history lag function value at the sth moment, so as to form an initial salinity alignment sequence based on the difference; Specifically, the construction of the initial salinity alignment sequence involves adjusting the time axis of the upstream salinity time series based on the immediate travel lag function. Specifically, the difference between the upstream salinity value at the sth moment in the upstream salinity time series and the value of the immediate travel lag function at the sth moment is calculated. This difference is used to map the upstream salinity time points according to the immediate lag relationship, initially aligning the upstream series with the reference time base of the downstream series on the time axis. The initial salinity alignment sequence formed through this mapping provides a preliminary reflection of the temporal lead or lag relationship between the upstream salinity and the downstream salinity.

[0060] If the initial salinity alignment value at the sth moment in the initial alignment sequence is ≥ the initial salinity alignment value at the s+1th moment, the initial salinity alignment sequence is optimized, including: The first constraint includes: the initial salinity alignment value at the sth moment < the initial salinity alignment value at the s+1th moment; The second constraint includes: calculating the square value of the initial salinity alignment value at the s-th moment after optimization and the initial salinity alignment value at the s-th moment, and calculating the sum of the square values ​​at each moment; and minimizing the sum of the square values; Specifically, if the initial salinity alignment value at time s in the initial salinity alignment sequence is greater than or equal to the initial salinity alignment value at time s+1, this indicates a temporal disorder in the timeline mapping. This disorder may be caused by time folding due to the dramatic fluctuations in the immediate travel time lag function. This violates the physical property of unidirectional time flow and can lead to erroneous temporal correlations in subsequent data fusion. Therefore, the initial sequence must be optimized to eliminate this non-physical temporal inversion.

[0061] Specifically, the first constraint requires that the initial salinity alignment value at time s be less than the initial salinity alignment value at time s+1, effectively forcing the sequence to be strictly increasing. This ensures that the optimized sequence adheres to the unidirectional nature of time, ensuring that the upstream salinity time mapping always progresses in the positive direction along the time axis, avoiding physical distortion caused by time reversal. This strictly increasing nature provides a consistent timing benchmark for subsequent fusion with downstream sequences, ensuring the physical plausibility of data alignment.

[0062] Specifically, the second constraint requires calculating the square of the optimized salinity alignment value at the sth moment and the initial salinity alignment value, then summing the squared values ​​across all moments and minimizing this sum. This ensures that the optimized sequence is as close to the initial sequence as possible while maintaining a strictly increasing order. Minimizing the sum of squares maximizes the preservation of salinity characteristics in the initial sequence, avoids loss of original data information due to overcorrection, and strikes a balance between physical plausibility and data fidelity.

[0063] An optimized salinity alignment sequence is formed based on the first constraint and the second constraint to obtain an aligned upstream salinity time series.

[0064] It should be noted that the synergistic effect of the first and second constraints ultimately produces the optimized salinity alignment sequence. The first constraint ensures the temporal rationality of the sequence, while the second constraint ensures the optimization result's fidelity to the original data. This optimization process eliminates temporal irregularities in the initial sequence while preserving the true dynamic characteristics of upstream salinity.

[0065] In one embodiment of the present invention, a fusion weight coefficient is determined according to a persistence parameter, and the upstream salinity time series and the downstream salinity time series are fused according to the fusion weight coefficient to obtain a fused salinity time series, including: Construct a time window at the sth moment to determine the duration parameter at the sth moment ; Specifically, a time window is constructed based on the sth moment, and the window length is the synodic period. The duration of the symbol of the instantaneous time delay function is counted through this time window, and the persistence parameter at the sth moment is calculated. The persistence parameter quantifies the long-term stability of the leading or lagging state of the upstream relative to the downstream at the sth moment, and its value range is A positive value indicates that the upstream is generally leading, and stability increases with increasing values. A negative value indicates that the upstream is generally lagging, and stability also increases with increasing absolute values. Using the persistence parameter as the basis for fusion weights ensures that the weight distribution is closely aligned with the physical lead-lag law, avoiding the subjective bias of empirical weights.

[0066] Construct the fusion weight coefficient with the duration parameter at the sth moment ; and limit the maximum value of the fusion weight coefficient to 1 and the minimum value to 0; In detail, the fusion weight coefficient is constructed based on the persistence parameter at the s-th moment. Specifically, the persistence parameter is converted into a weight value between 0 and 1 through linear mapping. When the persistence parameter is 1, the weight coefficient is 1, indicating that the upstream information is completely dominant in the fusion at this time; when the persistence parameter is -1, the weight coefficient is 0, indicating that the downstream information is completely dominant; the intermediate value of 0.5 distributes the weight proportionally. At the same time, the maximum value of the fusion weight coefficient is strictly limited to 1 and the minimum value is 0 to ensure that the weight is physically reasonable, that is, the sum of the contribution ratios of upstream and downstream information is always 1, to avoid distortion of the fusion result due to weight anomalies. In this way, the stability characteristics reflected by the persistence parameter are directly converted into an operational fusion weight.

[0067] Calculate the aligned upstream salinity value at time s and The first product of and the downstream salinity value at the sth moment and The second product of the first and second products is taken as the fused salinity value at the sth moment to form a fused salinity time series.

[0068] It should be noted that the calculation of the fused salinity value at the sth moment needs to be carried out in two steps. First, the first product of the aligned upstream salinity value at the sth moment and the fusion weight coefficient is calculated. The first product reflects the contribution ratio of the upstream information in the fusion result at the current moment; secondly, the second product of the downstream salinity value at the sth moment and 1 minus the fusion weight coefficient is calculated. The second product reflects the contribution ratio of the downstream information. The sum of the first product and the second product is the fused salinity value at the sth moment. Repeat this calculation process for all moments to finally form a fused salinity time series. This weighted fusion method enables the results to dynamically reflect the dominant relationship between upstream and downstream information. When the upstream leading stability is strong, the proportion of upstream information is high, and when the downstream dominant characteristics are obvious, the proportion of downstream information is high, thereby avoiding the physical dislocation caused by simple averaging or fixed weight fusion, and improving the accuracy of the fusion results.

[0069] In one embodiment of the present invention, discharge flow processing is performed based on the fused salinity time series, including: Determine the fused salinity value corresponding to the fused salinity time series at the current moment; If the fused salinity value is not within the preset salinity range, the flow of the upstream outflow port is controlled accordingly until the fused salinity value is within the preset salinity range.

[0070] Specifically, the preset salinity range is determined based on the actual water environment needs, ecological protection objectives, and watershed management standards of the monitoring area. For example, if the monitoring area is an estuarine spawning ground, the salinity range that ensures the survival and reproduction of key aquatic organisms (such as migratory fish) must be determined based on the appropriate salinity thresholds. If the monitoring area is near an industrial water intake, the range must be set based on the salinity requirements of industrial water use. This range must be verified through on-site research, ecological experiments, and industry standards before system deployment to ensure its scientific rationality and practical applicability.

[0071] Specifically, when the fused salinity value is higher than the upper limit of the preset salinity range, it indicates that there is a problem of excessive saltwater intrusion in the monitoring area. This may be due to the increased tidal power downstream, which causes the saltwater to advance upstream for an increased distance, or insufficient water from upstream, which cannot effectively dilute the saltwater. At this time, the system will execute a control instruction to increase the flow rate of the upstream spillway: by controlling the gate opening of the upstream spillway or the operating power of the pump group, the input of upstream freshwater is increased. When more freshwater enters the monitoring area, it will be fully mixed with the saltwater in the area, gradually reducing the overall salinity until the fused salinity value falls back to the preset range.

[0072] Specifically, when the integrated salinity value falls below the lower limit of the preset salinity range, it indicates excessive freshwater inflow into the monitored area. This could be due to increased rainfall upstream, resulting in a surge in water inflow, or excessive flow adjustments during previous discharge control. In this case, the system initiates a control command to reduce the flow at the upstream discharge outlet: by reducing the upstream gate opening or lowering the operating power of the pump unit, the upstream freshwater input is reduced. As the freshwater inflow decreases, saltwater from downstream is replenished upstream due to the tide, gradually raising the salinity in the area until the integrated salinity value returns to within the preset range.

[0073] Specifically, the intelligent control module continues to monitor changes in the integrated salinity value in real time. If the adjusted integrated salinity value still does not fall within the preset range, the flow adjustment range will be further adjusted based on the current deviation. Large deviations will result in an appropriate increase in the adjustment range, while small deviations will result in fine-tuning. If the adjusted integrated salinity value falls within the preset range, the system will stop flow adjustment and maintain the current flow state while continuing to monitor salinity changes to ensure long-term stability within a reasonable range, achieving dynamic and precise control of salinity in tidal river sections.

[0074] The above describes the embodiments of this embodiment, but this embodiment is not limited to the above specific implementation methods. The above specific implementation methods are merely illustrative and not restrictive. Ordinary technicians in this field can also make many forms based on the inspiration of this embodiment, all of which are protected by this embodiment.

Claims

1. An intelligent control system for water environment monitoring, characterized in that: include: The upstream monitoring unit is located at the upstream section near the tidal boundary and collects upstream salinity time series; The downstream monitoring unit is located at the downstream section near the tidal boundary and collects the downstream salinity time series; The signal processing module is connected to the upstream monitoring unit and the downstream monitoring unit and is configured to: The upstream salinity time series and the downstream salinity time series were filtered to separate the subtidal components. Based on the subtidal component, the upstream syzygy reference phase and the downstream syzygy reference phase are obtained by Hilbert transform. Calculate the cross-node and cross-layer syzygy phase difference based on the upstream syzygy reference phase and the downstream syzygy reference phase, and convert the syzygy phase difference into an instantaneous travel time delay function based on the syzygy beat frequency; The fusion processing module is connected to the signal processing module and is configured as follows: Taking the synodic period as the time window, the persistence parameter is calculated according to the symbol duration of the immediate travel time lag function. The upstream salinity time series is time-axis mapped based on the immediate travel time lag function to obtain the aligned upstream salinity time series. The fusion weight coefficient is determined according to the persistence parameter, and the aligned upstream salinity time series is fused with the downstream salinity time series according to the fusion weight coefficient to obtain the fused salinity time series. Intelligent control module that performs discharge processing based on the fused salinity time series.

2. The intelligent control system for water environment monitoring according to claim 1, characterized in that: The upstream and downstream salinity time series are filtered to separate the subtidal components, including: Based on the astronomical angular frequency set Determine the upper limit of the low-pass passband and the start of the stopband ;in, , represents the angular frequency of the M2 tidal component, represents the angular frequency of the S2 tidal component, represents the angular frequency of the K1 tidal component, represents the angular frequency of the O1 partial tide; Perform fast Fourier transform on the upstream salinity time series and the downstream salinity time series to obtain the upstream frequency domain spectrum and downstream frequency domain spectrum ; Constructing a zero-phase low-pass transfer function ;in, exist When the value is 1, The value is 0 when When the cosine smooth transition is obtained, the upstream low-pass main filter is calculated. and downstream low-pass main filter ;in, Belongs to the upstream frequency domain spectrum or the downstream frequency domain spectrum; right and Applying a narrowband notch transfer function ,as follows: in, Indicates the M2 tidal or S2 tidal notch bandwidth; The product of the narrowband notch transfer function and the zero-phase low-pass transfer function is taken as the total transfer function; The upstream subtidal component is obtained by performing an inverse fast Fourier transform on the product of the total transfer function and the upstream low-pass main filter; the downstream subtidal component is obtained by performing an inverse fast Fourier transform on the product of the total transfer function and the downstream low-pass main filter.

3. The intelligent control system for water environment monitoring according to claim 2, characterized in that: Based on the subtidal component, the upstream and downstream syzygy reference phases are obtained through Hilbert transform, including: The upstream subtidal component and the downstream subtidal component are removed from the mean and standardized to obtain the standard upstream subtidal component and standard downstream subtidal component ; Will and The absolute difference is taken as the syzygy beat frequency; Generates a time-varying reference waveform using the sine value of the syzygy beat frequency , and calculate the reference phase of the reference waveform through Hilbert transform ; The upstream analytical signals of the standard upstream subtidal component and the standard downstream subtidal component are constructed using the reference phase. and downstream analytical signals ;in, represents the Hilbert transform operation, i is the imaginary unit, ; Calculate the upstream instantaneous phase of the upstream analytical signal , calculate the downstream instantaneous phase of the downstream analytical signal ;in, represents the binary inverse tangent function, represents the imaginary part of the analytical signal, Real part of the analytical signal; Calculate upstream syzygy reference phase ; Calculate downstream syzygy reference phase .

4. The intelligent control system for water environment monitoring according to claim 3, characterized in that: The syzygy phase difference across nodes and layers is calculated based on the upstream syzygy reference phase and the downstream syzygy reference phase, and the syzygy phase difference is converted into an instantaneous travel time delay function based on the syzygy beat frequency, including: The phase difference between the upstream syzygy reference phase and the downstream reference phase is obtained by the wrap function to obtain the main value phase difference; wherein, the wrap function maps the phase difference value to the interval ; The main value phase difference is continuously expanded by unwrap to obtain the continuous phase difference; The ratio of the continuous phase difference to the syzygy beat frequency is used as the instantaneous time delay function .

5. The intelligent control system for water environment monitoring according to claim 4, characterized in that: Taking the synodic period as the time window, the persistence parameters are calculated according to the symbol duration of the instantaneous travel time lag function, including: The ratio of 2π to the synodic beat frequency is taken as the synodic period ; By time Defining a time window ; Through the time window Collect the instantaneous time-lag function value at fixed time intervals; If the value of the instantaneous time-history lag function is greater than 0, the sign is positive 1; If the value of the instantaneous time-lag function is less than 0, the sign is negative 1; If the value of the instantaneous time-lag function is equal to 0, the difference between the two adjacent instantaneous time-lag function values ​​on the time scale is calculated, and the corresponding sign is determined according to the positive or negative value of the difference; Determine the moment when the value of the instantaneous time-history lag function is positive 1 to obtain a first duration; determining a moment when the value of the instantaneous time-history lag function has a sign of negative 1 to obtain a second duration; The ratio of the difference between the first duration and the second duration to the synodic period is taken as the moment The persistence parameter.

6. The intelligent control system for water environment monitoring according to claim 5, characterized in that: The upstream salinity time series is mapped to the time axis based on the instantaneous travel lag function to obtain the aligned upstream salinity time series, including: Calculate the difference between the upstream salinity value at the sth moment in the upstream salinity time series and the instantaneous time history lag function value at the sth moment, so as to form an initial salinity alignment sequence based on the difference; If the initial salinity alignment value at the sth moment in the initial alignment sequence is ≥ the initial salinity alignment value at the s+1th moment, the initial salinity alignment sequence is optimized, including: The first constraint includes: the initial salinity alignment value at the sth moment < the initial salinity alignment value at the s+1th moment; The second constraint includes: calculating the square value of the initial salinity alignment value at the s-th moment after optimization and the initial salinity alignment value at the s-th moment, and calculating the sum of the square values ​​at each moment; and minimizing the sum of the square values; An optimized salinity alignment sequence is formed based on the first constraint and the second constraint to obtain an aligned upstream salinity time series.

7. The intelligent control system for water environment monitoring according to claim 6, characterized in that: The fusion weight coefficient is determined according to the persistence parameter, and the upstream salinity time series and the downstream salinity time series are fused according to the fusion weight coefficient to obtain the fused salinity time series, including: Construct a time window at the sth moment to determine the duration parameter at the sth moment ; Construct the fusion weight coefficient with the duration parameter at the sth moment ; and limit the maximum value of the fusion weight coefficient to 1 and the minimum value to 0; Calculate the aligned upstream salinity value at time s and The first product of and the downstream salinity value at the sth moment and The second product of the first and second products is taken as the fused salinity value at the sth moment to form a fused salinity time series.

8. The intelligent control system for water environment monitoring according to claim 7, characterized in that: Perform discharge processing based on the fused salinity time series, including: Determine the fused salinity value corresponding to the fused salinity time series at the current moment; If the fused salinity value is not within the preset salinity range, the flow of the upstream outflow port is controlled accordingly until the fused salinity value is within the preset salinity range.

Images